Download Some pyridoxal analogs and their transamination with amino acids

Document related concepts

Multi-state modeling of biomolecules wikipedia , lookup

Metabolic network modelling wikipedia , lookup

Enzyme wikipedia , lookup

Deoxyribozyme wikipedia , lookup

Evolution of metal ions in biological systems wikipedia , lookup

Fatty acid synthesis wikipedia , lookup

Basal metabolic rate wikipedia , lookup

Photosynthetic reaction centre wikipedia , lookup

Citric acid cycle wikipedia , lookup

Nucleic acid analogue wikipedia , lookup

Peptide synthesis wikipedia , lookup

Genetic code wikipedia , lookup

Hepoxilin wikipedia , lookup

Metabolism wikipedia , lookup

Metalloprotein wikipedia , lookup

Biochemistry wikipedia , lookup

Biosynthesis wikipedia , lookup

Amino acid synthesis wikipedia , lookup

Transcript
Retrospective Theses and Dissertations
1964
Some pyridoxal analogs and their transamination
with amino acids
Jerry David Albert
Iowa State University
Follow this and additional works at: http://lib.dr.iastate.edu/rtd
Part of the Biochemistry Commons
Recommended Citation
Albert, Jerry David, "Some pyridoxal analogs and their transamination with amino acids " (1964). Retrospective Theses and Dissertations.
Paper 3833.
This Dissertation is brought to you for free and open access by Digital Repository @ Iowa State University. It has been accepted for inclusion in
Retrospective Theses and Dissertations by an authorized administrator of Digital Repository @ Iowa State University. For more information, please
contact [email protected].
This dissertation has been
65-4589
microfilmed exactly as received
ALBERT, Jerry David,1937SOME PYRIDOXAL ANALOGS AND THEIR TRANS­
AMINATION WITH AMINO ACIDS.
Iowa State University of Science and Technology
Ph.D., 1964
Chemistry, biological
University Microfilms, Inc., Ann Arbor, Michigan
SŒIE PYRIDOXAL ANALOGS AND THEIR TRANSAMINATION
WITH AMINO ACIDS
by
Jerry David Albert
A Dissertation Submitted to the
Graduate Faculty in Partial Fulfillment of
The Requirements for the Degree of
DOCTOR OF PHILOSOPHY
Major Subject:
Biochemistry
Approved:
Signature was redacted for privacy.
In Charge of Major Work
Signature was redacted for privacy.
Keaa or major Department
Signature was redacted for privacy.
Daan of G<^duate College
Iowa State University
Of Science and Technology
Ames, Iowa
1964
11
TABLE OF CONTENTS
Page
INTRODUCTION
1
REVIEW OF PERTINENT LITERATURE
5
Vitamin Bg Chemistry and Function
5
Absorption Spectra and Ionic Equilibria of
Some Pyrldoxlne Analogs
6
Equilibria of Imlnes of Pyrldoxal Analogs
20
Function of Vitamin Bg In Transamination
47
Nonenzymlc Transamination In Model Systems
55
EXPERIMENTAL
74
Materials
74
Spectrophotometrlc Measurements of the Ionic
Equilibrium of Pyridoxal-N-methochloride
77
Spectrophotometrlc Determination of the For­
mation Constant for the Hydrogen-bonded
Aldlmine of Pyrldoxal-N-methochlorlde and
Valine
79
General Procedure for Nonenzymlc Transamination
in Model Systems
83
Analytical Methods
84
Kinetic Methods
97
RESULTS AND DISCUSSION
107
Absorption Spectrum and Ionic Equilibrium of
Pyridoxal-N-methochloride
107
Equilibrium Constant for the Formation of the
Hydrogen-bonded Aldlmine of Pyridoxal-N-metho­
chloride and Valine
116
Equilibrium Constants for the Formation of
Aldimlnes of Leucine and Deoxypyrldoxal,
Leucine and Pyrldoxal Phosphate as a
Function of pH
125
Ill
Page
Nonenzymic Transamination of Leucine with
Deoxypyrldoxal
131b
Nonenzymic Transamination of Leucine with
Pyridoxal Phosphate
I67
Intramolecular Acid-Base Catalysis of Non­
enzymic Transamination of Leucine with
Pyridoxal Phosphate
171
General Acid-Base Catalysis of Nonenzymic
Transaminations of Leucine with Pyridoxal
Analogs
l82
Metal Ion Catalysis of Nonenzymic Trans- ,
amination of Amino Acids with Pyridoxal
Analogs
210
Nonenzymic Transamination of Leucine with 5Carboxylate Analogs of Pyridoxal
228
Nonenzymic Transamination of Alanine Methyl
Ester with Pyridoxal Analogs
240
Deuterium Isotope Effect on the Rate-Limiting
Step in Nonenzymic Transamination
257
Future Pyridoxal Analogs Possibly Capable of
Enhanced Intramolecular Catalysis of Non­
enzymic Transamination
266
SUMMARY OP CONCLUSIONS
277
LITERATURE CITED
28l
ACKNOWLEDGEMENTS
297
DEDICATION
299
VITA
300
1
INTRODUCTION
One of the many functions of vitamin Bg, in its coenzyme
forms^ is the catalysis of the transamination of alpha-amino
acids and alpha-keto acids, a key reaction in nitrogen meta­
bolism. An object of this study has been to learn more about
the detailed action of vitamin Bg-dependent enzymes on amino
acids by inference from studies of nonenzymic reactions of
pyridoxal analogs and amino acids.
From previous studies of these nonenzymic reactions it is
known that the rates of both the formation of amino acid
imines of pyridoxal and the tautomeric interconversion of
these imines (the rate-determining step in transamination)
are several orders of magnitude lower than the corresponding
steps in enzymic catalysis.
Since general acid-base catalysis
of the reaction is Imown to occur in model transamination
systems, it is reasonable to ask whether certain catalytic
groups at the active site of a transaminase play a similar
role in enhancement of the enzymic reaction.
This situation seems likely because the available evidence
supports the catalytic role of acidic and basic groups at the
active sites of enzymes.
Such functional groups of amino
acids which might be available to enzymes for this purpose •
include the imldazolyl group of histidine, the phenolic group
of tyrosine, the guanidino group of arginine, the epsilonamino group of lysine, the hydroxyl groups of serine and
2
threonine, and the terminal carboxyl groups of aspartlc and
glutamic acids.
But the great efficiency of the enzymic catalysis is still
not fully understood; other factors must be considered.
Could
the dissociation constants of these acidic and basic groups be
important factors in effecting the efficiency of the catalysis?
If so, then how does the stereochemical arrangement of these
groups in relation to the coenzyme-substrate complex account
for the catalysis? Furthermore, how important are dynamic
conformational effects of the protein to the process of bring­
ing the catalytic groups and the coenzyme-substrate complex
into a critical configuration for polarizing the appropriate
chemical bonds required in the transamination of alpha-amino
acids and alpha-keto acids?
However, the problem to be dealt with here is how many of
these enzymic features of catalysis can be duplicated in a
model system?
The model transamination system of pyridoxal
(vitamin Bg aldehyde) and an amino acid or of pyridoxamine
(vitamin Bg amine) andia keto acid has been improved by adding
intermolecular catalysts such as metal ions and acid-base
buffers.
This model transamination system might be further
improved with Intramolecular catalysis by substituting certain
functional groups in the pyridoxal structure and by altering
the amino acid structure, such as by esterification, in order
to simulate effects of the enzyme on its substrate.
An approach which has added and could add further to our
3
I
understanding of the mechanisms of reactions on enzyme surfaces
Is the spectral measurement of pyrldoxal analogs and of their
reactions with amino acids.
These model systems are also
studied with the Intent of finding suitable conditions for con­
veniently comparing reactivities of pyrldoxal analogs and
measuring catalytic effects of the Incorporated substituents,
as well as of various Intermolecular catalysts.
Such exper­
imental conditions include those of temperature, pH, and con­
centration of reactants and catalysts.
Analytical methods, such as chemical, chromatographic, and
spectrophotometric techniques, and appropriate graphical
methods for analyzing equilibria and kinetic data need to be
adapted in order to measure these catalytic effects in these
nonenzymic reactions.
The transamination reaction is complex
in that Imine intermediates are involved in equilibria with
reactants and products, and the amounts and reactivities of
each of their ionic species are Important considerations in
determining their contributions to the overall reaction rates.
Measurements of, not only reaction rates, but also ionic
equilibria of each pyrldoxal analog and its amino acid Imines,
as well as the formation constants of these imines, need to be
made. Thus, in studying the transamination reaction in a
model system many factors need to be controlled or quantita­
tively taken into account in any accurate interpretation of
kinetic results.
In particular, Intermolecular catalytic effects of metal
4
ions, such as aluminum (III), copper (II), and zinc (II), of
acid-base buffers, such as acetate, phosphate, and imidazole,
are to be measured.
Intramolecular catalytic effects of a
positive charge on the pyridine nitrogen atom of pyridoxal and
of substituent groups at the 5-position of pyridoxal are to be
determined. The latter substituents include hydroxymethyl,
methylol phosphate, propionate and carboxylate groups, with
reference to a methyl group.
The deuterium isotope effect in
deuterio-leucine and the effect of an esterified amino acid
on the rates of these nonenzymic transamination reactions are
to be measured.
The purpose of these investigations is to add to the
understanding of the mechanism of pyridoxal-mediated trans­
amination of amino acids and of the increased reactivity of
pyridoxal phosphate in the enzymic reactions, which possibly
is due to the nature of catalytic groups in the protein at the
active site and to the mode of attachment of the coenzyme to
apoenzyme.
It is further hoped that this experimental approach
will lead to the design and synthesis of pyridoxal analogs
capable of carrying out the intramolecular acid-base catalysis
of the intermediate imines with amino acids more efficiently
than the presently available analogs, and perhaps as efficient­
ly as the enzymes found in nature. Such an investigation of
the chemical action of vitamin B, coenzyme (or of pyridoxal
analogs) constitutes an important part of the basic research
on a fundamental problem of biochemistry, namely, the elucida­
tion of the mechanisms of enzymic reactions.
5
REVIEW OF PERTINENT LITERATURE
Vitamin
6
Chemistry and Function
Thirty years ago vitamin Bg was discovered and defined as
a discrete factor in the vitamin B-complex (Gyorgy, 1934).
During the subsequent three decades a wealth of information
has been gained about the chemistry and biological role of
this vitamin.
This information has pointed to the central role
of vitamin Bg in nitrogen metabolism.
In his review entitled
"Pyriaoxal Phosphate," a vitamin B, coenzyme, Braunstein (i960,
p. 180), the discoverer of biological transamination, concludes:
"Pyridoxal catalysis is clearly one of the most important and
versatile among the fundamental enzymic mechanisms constituting
the chemical basis of life.
Its very ancient origin in the
evolution of living matter is evident from the many strategic
functions of pyridoxal phosphate enzymes in the biosynthesis
and transformations of virtually all biologically important
nitrogen compounds."
A recent symposium on the Chemical and
Biological Aspects of Pyridoxal Catalysis has reviewed the
current knowledge and
has indicated the directions of research
in this field (Snell et
, 1963).
Another review on the
chemistry of the vitamin itself (pyridoxine) has recently
appeared (Wagner and Polkers, 1964).
In its coenzyme forms vitamin Bg participates in a variety
of important enzyme-catalyzed reactions (Braunstein, 196O;
6
Gonnard, 1962; Snell, 1958, 1962).
Our present understanding
of the roles of vitamin Bg 'coenzymes in enzymic reactions has
developed from several different types of experimental ap­
proaches. The approaches which will be emphasized in this
review include investigations on nonenzymic model systems and
studies of enzyme systems reacting with antagonists (carbonyl
reagents and structural analogs) of vitamin Bg.
Absorption Spectra and Ionic Equilibria
of Some Pyridoxine Analogs
The electronic absorption spectra of pyridoxine and other
hydroxypyridine derivatives change markedly with pH when
dissociation of the phenolic or pyridinium groups occurs.
These spectra and the ionic equilibria of these compounds in
aqueous solution have been compared and correlated with those
of a number of other pyridine and benzene derivatives.
Effects
of molecular structure on these spectra and on the observed
ionization constants have been discussed (Williams and Neilands,
1954; Metzler and Snell, 1955; Nakamoto and Martell, 1959 a/b).,
A summary of the current knowledge in this area of vitamin Bg
chemistry has been presented in a recent review (Martell,
1963).
Fyridoxol, PO-Figure 1
The spectrum of pyridoxol may be compared with that of 3hydroxypyridine.
The absorption maximum at 291 mp. for an acid
solution of pyridoxol has been assigned to the cationic form
7
(as a-Plgure 2). In neutral solution the absorption maximum
at 324 m;a has been assigned to the dipolar Ionic form (as bFlgure 2), in which the phenolic group has dissociated.
Spec-
trophotometrlc and potentlometrlc titration data have been
used to calculate the ionization constants for the phenolic
and pyrldlnium groups of pyridoxine analogs.
The first
apparent constant has been observed in the range of pK 3.35.0 (pK^-Table 1) for these analogs. The second apparent
constant has been observed in the range of pK 6.8-9.0 (pK^Table l), and the absorption maximum at 310 vnp. for a basic
solution of pyrldoxol has been assigned to the anionic form
(as c-Pigure 2).
5-Deoxypyridoxal (DPL) and 3-hydroxypyrldlne4-aldehyde (HPA)-Flgure 1
The introduction of an aldehyde substituent in the 3hydroxypyrldlne nucleus results in changes both in absorption
spectrum and in ionization constants.
In general, both
apparent ionization constants decrease somewhat from their
values in pyrldoxol. The electronic absorption bands have
been assigned to two sets of ionic structures.
Those for the
hydrated aldehydes (as C-Plgure 2) correspond with those of
pyrldoxol (alcohol structures).
The unhydrated aldehyde forms
(as B-Flgure 2) of deoxypyrldoxal have absorption maxima at
342 rap. for the cation, 383 mja for the dipolar ion, and 391 mp
for the anion, which correspond closely with the absorption
bands for 3-hydroxypyridlne-4-aldehyde.
Figure 1. Structures and nomenclature of some pyridoxal ana­
logs
(PO)
3-Hydroxy-2-methyl-4j5-dihydroxymethylpyridlne: pyridoxol, vitamin Bg alcohol
Structure with R = R' = H
(HPA) 3-Hydroxy-4-formylpyridine or 3-hydroxypyridine-4aldehyde
Structures with R = methyl and with 5-substituents
(PL)
R' = hydroxymethyl:
Pyridoxal, vitamin Bg aldehyde
(DPL) R' = methyl: 5-Deoxypyridoxal, a vitamin Bg anti­
metabolite
(PLP) R' = methylol phosphoric acid ester: Pyridoxal-5-phosphate, a vitamin 3g coenzyme
(CPL) R' = carboxy: ,"5-Pyridoxalylic acid" or "5-carboxypyridoxal
(FPL
R' = carboxymethyl:
or PLF)
(APL
R' = carboxyethyl:
or PLA)
"alpha^-Pyridoxalylformic acid"
"alpha^-Pyridoxalylacetic acid"
Structures with an N-methyl substituent
(PLM) R' = hydroxymethyl;
Pyrldoxal-N-methochloride
(MPLP) R' = methylol phosphoric acid ester: 5-Phosphopyridoxal-N-methochloride or N-methopyrldoxal-5-phosphate
(DPLM) R' = methyl: 5-Deoxypyridoxal-N-methochloride
Pyridoxamine analogs
(PM)
3-hydroxy-2-methyl-4-aminomethyl-5-hydroxymethylpyrldine:
Pyridoxamine, vitamin Eg amine
(PMP) Pyridoxamlne-5-phosphate, a vitamin 3g coenzyme
(DPM) 5-Deoxypyrldoxamlne, a vitamin Bg antimetabolite
9
(PO) Pyridoxol
(PM) Pyridoxamlne
HO
OH
R = R' = H
(HPA) 3-Hydroxypyrldlne-4-aldehyde (PLM) Pyridoxal-Nmethochlorlde
R = -CH
(PL)
R'
-CHgOH, Pyridoxal
(DPL) R'
-GH^, Deoxypyridoxal
(PLP) R'
-CHgOPO^Hg, Pyridoxal phosphate
(CPL) R'
-CO H, "5-Pyridoxalylic acid" or "5-carboxypyridoxal" (3-hydroxy-2-methyl-5-carboxypyridine-4-aldehyde)
(PPL) R' = -CH 00 H, "alpha^-Pyridoxalylformic acid" (32
hydroxy-2-methyl-5-carboxymethylpyridine-4-aldehyde)
(APL) R' = -(CH ) CO H, "alpha^-Pyridoxalylacetic acid" (32
hydroxy-2-methyl-5-carboxyethylpyridlne-4-aldehyde)
Figure 2.
Aqueous solution equilibria of ionic forms of pyridoxal' analogs in
addition to those for deoxypyridoxal reported by Metzler and Snell
(1955) and for hydroxypyridine aldehydes reviewed by Martell (1963)
CBTION
H
DIPOLFIR ION
HNION
DmNlONI
0
INTERWRL
HEMiRCETRL
F1
I •
pKb
UNlHyORRTED HO
RLDEH/DE
pib
©
pK d
B
QUINOID
HyORRTED HO
RLDEHyOE
C
"H
Table 1. Ionization constants and assignments of observed electronic absorption
bands with mole.r absorbancy indices of ionic forms of some pyridoxine
analogs at Sgoc, 0.1 M ionic strength^
B
PO^ d K 5.00
a
8.97
C
D
a
291
E
8.6
pK^
CD
5.00
b
(232)8 2.1
E
pK^
324
7.2 8.97
254
3.9
C
c
D
E
pK
C
310 6.8
245
6.3
•
PL° pK
a
PKb
PKc
pKa
pKg
4.3
5.3
8.6
7.6
1.06
Aa 288
4.20
9.0
(230)8 3.3 +0.02
Ab 317
252
Bb 390
8.9 8.66
5.8 +0.06
0.14
Ac 302 5.71
240 8.4
Be 390 0.2
A^B^ = 80
A/B^ = 2.8
Ayc^ = 120
Ayc^ = 500
Cb/E^ = 0.66
13
Ad^
= 0.006
^Throughout this table the following letters are substituted for these headings
A = PL analog (Figure l); B = Microscopic (molecular) constants (except for phosphoryl and amino pK's which do not affect spectra): C = Ionic form (Figure 2); D =
Wavelength of maximum absorbancy, mu (s = shoulder); E «• Molar absorbancy index, a
X 10-3
"Source: Lunn and Morton, 1952; Williams and Neilands, 1954; Peterson and
Sober, 1954; Metzler and Snell, 1955; Martell, 1963
^Source: von Viscontini e;t
, 1951; Lunn and Morton, 1952; Williams and
Neilands, 1954; Peterson and Sober, 1954; Metzler and Snell, 1955; Nakamoto and
Martell, 1959b; Martell, 1963
^Ad has a^jj - 5.8 X 103 at 300 mp
V
Table 1.
(continued)
D
A
E
PKn
D
E
Pm® pK 4.05
^
Aa 293.5 7.22 4.05
(220)5
+0.03
Ab 323
7.80
254.5
4.8
Bb 400
0.17
Ab/Bb = 46
HPA
Ba 324
0.3
(220)8 4.2
Ca 285
6.2
Bb
Cb 324
240
pK
4.2
pKg 6.6
PK^ 6.4
pKg 0.22
4.05
DPL® pK^ 3.8 Ba 342
1.9
4.17
(255
)8
1.6
+0.04
pK^ 4.5 Ca 294.5 6.4
PKb 4.3
= 3.4
pKg 7.8
pKd 8 . 0
pK^ 0 . 2 2
2.7
5.8
pKr
D
6.77 Bc 384
Ce (245)8
E
pK,
i:§
Bb 383
4.26 8.14 Bc 391
6.2
(250)8 4.2 4-0.08
(265)8 3.4
Cb 324
2.6 0.6
Ce 306
cyB = 0.66
C /B = 0.06
c c
^Source: Johnston et
, 1963
Source: Nakamoto and Martell, 1959bj î-Tartell, 1963
SSource: Heyl et
, 1953; Metzler and Snell, 1955; Nakamoto and Martell,
1959b; Martell, 1963
Table 1.
(continued)
B
PLph pK^, 5.9
D
E
D
E
pK
D
3.44
Bb 384
5.00 8.45 Be 388
4
.00
388
.90 +0.04 (270)s
4
pKg, 7.96
2.56 -3.4 Co 305
4.14
325
phosphoryl Cb 330
2.50 8.69
Ca 295 6.7
1.4
6.2 (250)3 4.00
+0.4
+0.2
Ca/Ba = CDVBb - C c/Bc = 0.36 +0.07
—
Ba 340
APL^
8.4 Be 390
3.96
Bb 390
Cc
Ca 290
Cb
PM^
a 293 8 . 5
3.54
b 325
7.70 8.21 0 308
4.60 7.90
3.31
253
245
Ba 335
1.8
PK^
E
pK .
3
C
6.6
3.0
1.1
7.30
5 .90
amino pK
10.63
10 .4
^Source; Umbreit jet a^., 1948; Heyl e;b a^. 1951; Lunn and Morton, 1952;
Peterson .et a^., 1953; Williams and Neilands, 1954; Peterson and Sober, 1954; Metz1er and Snell, 1955; Christensen, 1958; Bonavita and Scardi, 1959; I^fe-rtell, 1963;
Anderson and Kartell, 1964
^Source: Tomita and Metzler, 1964; F. Scott Furbish, Ames, Iowa, Iowa State
University of Science and Technology, Department of Biochemistry and Biophysics.
1964. Spectrophotometric determination of ionization constants of "alpha5-pyridoxalylacetic acid." Private communication.
JLunn and Morton, 1952; Williams and Neilands, 1954; Metzler and Snell, 1955
Table 1.
A
PMpk
B
(continued)
C
D
a 293
E
9.00
p K ^
C
D
E
p K g
3.69
b 325 8.3 5.76
3.25
253 4.7
+0.07
phosphoryl
pK<2.5
C
D
E
c 308 8.0
245 6.7
amino
pK 10.92
^Source:
Williams and Neilands, 1954; Metzler and Snell, 1955
p K
C
l6
The relative ratios of amounts of hydrated to unhydrated
aldehydes have been estimated from the molar absorbancy indices
(C/B-Table l).
The decrease in the relative concentration of
hydrated aldehyde with increasing pH is due to a decrease in
electron-withdrawing power of the pyridine ring as a result of
the step-wise dissociation of protons.
In decreasing the
positive character of the charge on the formyl carbon atom,
the addition of water across this carbonyl bond will be less
likely.
Pyridoxal-5-phosphate (PLP) - Figure 1
The spectrum of the vitamin Bg coenzyme, pyridoxal phos­
phate, is almost identical with that of deoxypyridoxal, a
Potent vitamin B, antimetabolite, over the whole pH range (von
Viscontini et al., 1951; Heyl
al^., 1953), and the results
of the interpretation of the spectrum of the latter can also
be applied to that of the coenzyme (Metzler and Snell, 1955).
Ionizations of the phosphoric acid ester group (phosphoryl pK's
Table l) cause little difference in the spectrum of pyridoxal
phosphate from that of deoxypyridoxal (Williams and Neilands,
1954).
Recently, molecular species of pyridoxal phosphate have
been studied in DgO solution by infrared spectra at various pD
values, and the results have been compared with those obtained,
in the same manner, from pyridoxal, pyridoxamine, pyridine,
and hydroxypyridine derivatives (Martell, 1963; Anderson and
17
Martellj 1964).
Spectral band assignments have been made and
pK' values have been obtained. The ionization constants have
a
been used to calculate some microscopic (molecular) equilibrium
constants for pyridoxal phosphate in aqueous solution.
These
pK^ values (Figure 3) are for individual group ionizations, in
contrast to the apparent pK' values (Table 1) reported prea
viously for pyridoxal phosphate. The apparent pK'g^ values
agreed reasonably well with the previously reported values
except that pK^ (Figure 2, Table l) for dissociation of the
phenolic group was 3.44, in contrast to the earlier values of
4.14 and 4.0 (Christensen, 1958; Williams and Neilands, 1954).
Also, the ratio of hydrated to unhydrated aldehyde forms
was constant, within experimental error (C/B-Table 1), over
all pD values from 4.7 to 11.0, in contrast to the increasing
proportion of the unhydrated aldehyde form in deoxypyridoxal
and pyridoxal, with increasing pH (Metzler and Snell, 1955).
The pK's also were reported for two different ionic strengths
(O.l and 2.0 M), and an appreciable difference was noted
(Anderson and Martell, 1964).
Pyridoxal (PL) - Figure 1
The introduction of a hydroxymethyl substituent at the
5-position of 3-hydroxypyridine-4-aldehyde causes considerable
changes in the spectrum due to the predominance of the cyclic .
hemiacetal form (A-Figure 2). The hemiacetal forms predominate
Figure 3.
Aqueous solution equilibria of some ionic forms of unhydrated
pyridoxal phosphate showing microscopic or molecular dissociation
constants (pKa's); from Infrared spectra of DoO solutions
(Anderson and Martell, 1964; Kartell, 1963)
OaPO
0pK5,9±Q!fpKoj)
• WL
%
pKqô9
pK Q20
e.
'HO.
'3PO
r^96(pKa)
±qo5
±Q05
ae
D3PO
pK 5,2±QI
^^OjPO
pKTTTiqi
20
over the free aldehyde forms over the whole pH range because
of the stability of the 5-membered ring (Metzler and Snell,
1955; Nakamoto and Kartell, 1959t>). However, the ratio of
hydrated to unhydrated aldehyde (C/B-Table l) is about the
same as that for deoxypyridoxal at similar pH values
(Metzler and Snell, 1955; Table l). In general, the elec­
tronic absorption spectrum of pyridoxal is similar to that of
pyridoxol, except for the small absorption bands due to the
free aldehyde forms (Table l).
Pyridoxal-N-methochloride (PLM) - Figure 1
The absorption spectra and acid dissociation constants
(pK 's) of the phenolic groups of pyridoxal and pyridoxal-N&
methochloride have been compared (Johnston et al., 1 9 6 3 ) .
Like the pyridoxal hemiacetal cation (Aa-Pigure 2), pyridoxal
methochloride can dissociate to form the dipolar ion (AbPigure 2).
However, unlike pyridoxal, the N-methyl analog can
not form an anion (Ac-Figure 2),, and this was reflected in the
same absorption spectrum at high pH as at neutral pH.
A
small amount of free aldehyde should also be present in
equilibrium with the hemiacetal form, as in the case of
pyridoxal. (See Results and Discussion for further details
of this study.)
Equilibria of Imines of Pyridoxal Analogs
"The Bg vitamins, and coenzymes, their imines (Schiff
bases), and the corresponding metal chelates exist in many
21
equilibrium states, including various tautomers in neutral and
ionic forms.
Considerable information is currently available
on the structure, absorption spectra and ionization and
stability constants of these forms and on the effects of pH
and other factors on.their equilibria.
These data provided a
basis for successful application of spectrophotometric and
other methods in investigations on the linkage and structure
of the coenzyme moiety of highly purified Bg enzymes in aqueous
solutions at different hydrogen ion-concentrations and in the
presence of specific substrates or inhibitors" (Braunstein,
i960, p. 182).
Absorption spectra and ionic equilibria of imines
When solutions of pyridoxal or pyridoxal phosphate and an
amino acid are mixed characteristic electronic absorption
spectral changes result from rapid formation of aldimines
(Eichorn and Dawes, 195^; Metzler, 1957; Matsuo, 1957a;
Christensen, 1958; Cattaneo et al., 196O; Lucas _et^., 1962j
Heinert and Martell, 1963a). A recent review has summarized
and interpreted the spectral changes associated with the
formation of various ionic species of these aldimines (Martell,
1963).
In general, the hydrogen-bonded aldimine forms of pyridoxylidene amino acids have been compared with those of salicylaldehyde imines.
The 404-2$ mju absorption (TT- TT^*
transition) and the 265-80 mp absorption (TT- TTg* transition)
bands have been assigned to the hydrogen-bonded aldimine
Figure 4. General Ionic equilibria of pyridoxylidene amino acids
DIPOLAR OTTION
DIPOL-RR ION
^OH
R^^<L©
Y<p
J,
/
DIPOLBR "MNION
pKg
J,
y
DIBNION
,
pr\3
KETOENFIMINE
Ra H
Fib H
FIMIDE VINYLOG
Cb H
METAL CHELATE
DCL
Eb H
Be
TIMIDE VIN/L06
ENOLIMINE
HEMIBCETRL
OR
HYDRRTED
RLDIMINE
Fc-cL
CH3
Table 2.
Ionization constants and assignments of observed electronic absorption
bands of ionic forms of some pyridoxylidene amino acids at 25°C
Ionic forms (Figure 4)
Aa-c
Imine
PL-Val®-
2 ^Max
mja
5.9
4l4
280
Ad
a~
xlO"3
pK'^ Amax
mjn
2+
Be
a~
Xmax
xlO~3
6.24 10.49 367-70 7.88
7.23 +0.03
Cb-Zn
mja
Amax
316-30
238-60
377
277
a
pK'
xlO"3
8 .7 6.5
4.8
Da or Eb
Amax
mp
320-30
250-60
pK'l (2.3)^
PLP-Val° (6.3)
415
280
(10.8)
(11.5) 367
278-85
DPL-Leu^ (6.1)
PLM-Val^
8.0
4l4 (9)
294
Aa-b
419
7.5
288
4.7
234
(217)
Pc-d
378
323
8.4
383
7.4
280 5.1
Metzler, 1957; Davis et al., 1961; Heinert and Martell, 1963a;
^Source:
Martell, 1963
^Throughout, niombers in ( ) = value uncertain
5Source: Christensen, 1958
Source: RommelErika, and David E. Metzler, Ames, Iowa, Iowa State Univer­
sity of Science and Technology, Department of Biochemistry and Biophysics. 1964.
Spectrophotometrlc determination of formation constants of imines of leucine and
deoxypyridoxal. Private communication
^Source: Johnston et al., 1963
25
(Metzler, 1957; Lucas et al., 1962; Helnert and i^lartell, 1963a).
Protonations of the pyridine or imine nitrogen atoms are not
expected to result in more than a slight shift in these wave­
lengths of maximum absorbancy (Martell, 1963).
From Infrared
spectral studies it was found that the ketoenamine form
(A-Pigure 4), in which the proton of the hydrogen bond is
closer to the imine nitrogen atom than to the phenolic oxygen
atom, predominates over the enolimine form (Bc-Figure 4),
the other tautomer, in aqueous solution (Helnert and Martell,
1962).
The enolimine tautomer (Bc-Pigure 4) has absorption
bands similar to the hemlacetal (Eb) or hydrated imines (DaFigure 4):
316-30 mp. for TT^-transitions and 238-6O mp for
n^-transltions (Table 2).
Removal of the hydrogen-bonded proton, forming a dianion
(Ad-Figure 4), does result in a shift of the absorption maxi­
mum to lower wavelengths (Table 2), as does replacement of
I
this proton with a metal ion (Cb-Flgure 4; Table 2; Eichorn
and Dawes, 1954). The ionic equilibria of these imlne-metal
chelates have been studied spectrophotometrically. The
effect of chelation by metal ions on the pK'
of the pyri-
dinium group depended on the nature of the metal ion.
Values
have been reported for pK'^ from about 5.5 to 7.9 (Chrlstensen, 1959; Davis _et al., 1961).
Ionic equilibria of N-(3-hydroxy-4-pyridylmethylene)valine were studied quantitatively as a model system for
pyridoxal imines (Helnert and Martell, 1963b).
The ionization
26
constants are reversed in going from pyridoxal to its imine
forms.
The pK'^ of 8.6 for the pyridinium group in pyridoxal
changes to 5.9 In the imine, and the pK'^ of 4.2 for the
phenolic group in pyridoxal changes to 10.5 in the imine, as
a result of the formation of a rather stable hydrogen bond
between the phenolic oxygen atom, the imine nitrogen atom,
and perhaps the carboxylate group of the amino acid residue
(Ac-Figure 4; Metzler, 1957; Christensen, 1958).
In addition,
the ionic equilibria of ketimines from pyridoxamine and keto
acids have been determined spectrophotometrically (Figure 5;
Banks et al., 1961).
The elimination of the absorption band at 278 mp upon the
addition of acid or base to a neutral solution of pyridoxal
phosphate and valine was followed to find apparent pK values
of 6.3 and 11.5^ respectively.
Also, a pK value of 11.5 was
found for the elimination of the absorption band at 415 mp
upon the addition of alkali.
This was compared with a pK
value of 10,8 from a potentiometric titration, but the ionic
strength was 0.42 M, whereas it was only 0.10 M in the spectrophotometric titrations.
Because tri-and tetravalent ions
are involved in equilibria of pyridoxal phosphate imines, high
sensitivity to ionic strength was anticipated (Christensen,
1958).
The existence of imines of pyridoxal and amino acids (and
their metal chelates) has been proven by their isolation,
identification, structure elucidation, and syntheses, which
Figure 5. Absorption spectra and ionic equilibria of some ionic forms of
ketimines of pyridoxamine and pyruvate (R = CH )(Banks et al.,
1961)
3
DIPOLFIR CRT ION
DIPOLFIR ION
.0
©""M
6® pKk,l (2,5)
DIPOLFIR "ANION
R.
HO
c
L
J,
HO
©^M"
HO
Ô® pKttS 'Q3
J)B pK'„ 2 6,9
•CH3
(28^
HQ
Ht
DIFINION
29
i
have been accomplished repeatedly in a number of different
laboratories in the past twenty years.
Some of the more recent
accomplishments in this area also include reviews of previous
findings (Christensen, 1958, 1959; Pasella _et
, 1958;
Davis et aJ.J I96I; Heinert and Martell, 1962, 1963c; Cennamo,
1963).
Furthermore, the stereochemistry of the manganese
(ll)-pyridoxylidene valine chelate has recently been estab­
lished by x-ray diffraction (Willstadter et
1963),
which indicated an essential planarlty of the fused ring
systems (from the measured bond angles and lengths) and a
strong phenolic oxygen anion-metal cation bonding.
Mechanism of imlne formation
The mechanism of imlne formation has been described as
initiating with a nucleophilic attack of the uncharged amino
acid nitrogen atom on the electrophlllc carbonyl carbon atom
of the 4-formyl group of pyridoxal. Tautomerization of the
dipolar ion intermediate followed by dehydration of the
carbinolamine (tetrahedral addition intermediate) leads to
the imlne, a readily observed and a relatively more stable
intermediate in the reaction between pyridoxal and an amino
acid.
All of these steps leading to the imlne are equilibria
which are rate-determining at certain pH values (Figure 6).
This general mechanism of imlne formation and hydrolysis has
recently been substantiated with kinetic data (Cordes and
Jencks, 1962a, 1963; Martin, 1964), which have been reviewed
Figure 6.
Mechanism of imine formation; schematic comparison of aldimine for­
mation (Snell, 1962)
A. Prom an amino acid and pyridoxal
B. Prom an amino acid and a preformed aldimine of pyridoxal by
transaldimination
Step 1.
Nucleophllic attack of amino-N on carbonyl-C
(For semicarbazone formation: rate-determining for pH less than 7)
Step 2.
Tautomerization
Step 3. Dehydration
(For semicarbazone formation: ratedetermining for pH above 7 and general
• acid-catalyzed for pH less than 4)
(Jencks and Cordes, 19^3; Martin, 1964)
COo®
T
R
CO2®
—
R^COg®
^
ipr"^oH
CHRBINOLRMINE
R
CCfe®
.%yX =
T
j|
+H^0
El
RLDIMINE
u>
J
NHg
^
•RLDIMIÎ^E
HO
(
32
recently (Jencks and Cordes, 1963; Jencks, 1963; Martin, 1964).
It is interesting that the dehydration step is subject to
general acid catalysis (Cordes and Jencks, 1963; Jencks and
Cordes, 1963; Jencks, 1963).
Rates of imine formation
Formation and hydrolysis of imines occur rapidly and
reversibly in aqueous solution at room temperature.
A non­
aqueous medium, such as ethanol, enhanced the formation of
imines of pyridoxal or pyridoxal phosphate and amino acids
(Matsuo, 1957b).
In the system of pyridoxal and valine equi­
librium was achieved within ten minutes (Metzler, 1957).
Several investigators have reported absorbancy changes with
half-lives of a few minutes when pyridoxal and certain amines
were mixed or when pH values of the resulting solutions were
changed (Williams and Neilands, 1954; Metzler, 1957; Christensen, 1958; Fleck and Alberty, 1962).
Rate constants for the formation of carbinol amine (the
tetrahedral addition intermediate - Figure 6) and aldimine
were calculated from spectral changes in solutions of pyri­
doxal and alanine (Fleck and Alberty, 1962).
Depending upon
the conditions of pH and concentration, the former was 0.11
HM0.04/sec. at a maximum (half-time of about 0.7 sec.) and the
latter was from 0.003 to 0.065/sec. (half-times of 10-200 sec.)
at 25®C. From the rate constant reported for the hydrolysis
of the aldimine of pyridoxal and gamma-aminobutyrate (Olivo £t
al., 1963), the half-time can be estimated as five minutes.
33
From the observed rate and equilibrium constants reported for
formation of imines of pyridine-4-aldehyde and eleven amino
acids (Bruice and French, 1964), half-times could be calcu­
lated.
They ranged from about 47 seconds for leucine to I87
seconds for phenylglycine. From a similar quantitative anal­
ysis of the formation of imine of pyridoxal phosphate and
cysteine (and its S-derivatives) (Mackay, 1962), half-times
of 40 to 230 seconds, depending upon the conditions of pH
and concentration, can be calculated.
A slow imine-cupric
chelate formation was observed over a period of 24 hours to
equilibrium (not transamination) (Christensen, 1959). Other
metal ions, such as zinc and sodium, replace the proton in
hydrogen-bonded aldimines about as fast as imine formation
itself (Davis _e;t
, 1961).
The reactive form of pyridoxal for imine formation re­
quires the free formyl group, and the limiting factor in the
rate of imine formation from pyridoxal is the unfavorable
equilibrium between the hemiacetal and free aldehyde forms,
which also affects the final imine concentration or equi­
librium constant (Snell, 1958). The higher rate of semicarbazone formation from pyridoxal phosphate, compared to
pyridoxal, is the result of pyridoxal existing in aqueous
solution principally as the tinreactive internal hemiacetal
(Cordes and Jencks, 1962b; Jencks and Cordes, I963).
The
half-time for semlcarbazone formation from pyridoxal is about
3i minutes (k^^^ = 0.2/min.).
Also, imine from gamma-amino-
34
butyrate and pyridoxal phosphate forms almost Immediately, as
compared to a slightly slower Imlne formation from pyridoxal
(Ollvo et^., 1963).
Equilibrium constants for formation of Imlnes
Extent of Imlne formation depends upon pH, temperature,
nature of the solvent and presence of metal Ions.
More Imlne
Is formed from pyridoxal and gamma-amlnobutyrate with In­
creasing temperature (Ollvo ^
, 1963).
Imlne formation
Is also favored by non-polar solvents, such as ethanol (Matsuo,
1957b), and by metal Ions which form chelates (Longenecker and
Snell, 1957; Davis et aJ., 196I; Cennamo, 1963).
The equilibrium constants for formation of Imlne anions
from pyridoxal and amino acid anions approach maximum values
in the alkaline pH range.
Calculated values from spectral
data at 25°C agreed with a theoretical curve of pH versus the
logarithms of the equilibrium constants for formation of imlne
anion (log K^) from pyridoxal and valine (Metzler, 1957).
This theoretical curve was based on pK'^ values of pyridoxal,
amino acid and imlne, and is compared with similar curves for
imlnes of leucine with deoxypyridoxal and with pyridoxal
phosphate (Figure 15; Results and Discussion).
In addition, formation constants of ketimines of pyrldoxamine and pyruvate have been determined spectrophotometrlcally as a function of pH.
These experimental values agreed
with predicted formation constants calculated from pK'^ values
35
of the dissociable ketimines and reactants (Banks _et
, 1961).
A maximum formation constant of 9.8 was obtained at about pHlO,
and this may be compared with the formation constant of 9
obtained for the aldimine anion of pyridoxal and alanine
(Metzler, 1957).
The imine formation constant depends also somewhat on the
structures of the amino acid and pyridoxal analog. Leucine
was shown to have a favorable imine formation constant with
pyridoxal, compared to other amino acids, and its value was
about three times greater than that for alanine (Metzler,
1957), although reasons for this difference are unknown.
Equilibrium constants for imine formation from eleven amino
acids with pyridine-4-aldehyde were compared recently (Bruice
and French, 1964). In the graphical determination of these
constants between pH 7 and 10 it was concluded that only the
free amino group is essential to the over-all equilibrium.
The highest value (leucine) was three times that for the lowest
(phenylglycine). These values are between ten and thirty times
higher than those reported for pyridoxal and amino acids, due
to hemiacetal formation in the case of pyridoxal.
Also, imine
formation takes place to a lesser extent from pyridoxal and
amino acid esters than from pyridoxal, metal ions and amino
acids (Cennamo, I963, 1964).
Early imine formation constants for pyridoxal phosphate
and amino acids ranged from 10^ to 10^ in neutral, aqueous
solution, but the calculations from spectral data at 278 mp
36
took Into account the water (55.5 M) formed along with Imlne
(Matsuo, i957a). The formation constants for the.sum of
products from pyridoxal phosphate reacting with amino acids,
peptides and proteins were estimated from both decreasing
and increasing spectral changes (Christensen, 1958). These
values ranged from 15O to 8000 at pH 7-5 (Christensen, 1958).
More recently, the imlne formation constant for pyridoxal
phosphate and S-methylcysteine was 52/M at pH 6.3 and 450/M
at pH 7.9, 21®C (Mackay, 1962).
The values for pyridoxal
phosphate and gamma-aminobutyrate were 15.5/M at pH 6.3 and
lll/M at pH 7.4, 25°C, and were 25 times greater than those
for pyridoxal and gamma-aminobutyrate under the same con­
ditions (01ivo_e^^., 1963).
In summary, equilibrium constants for imine formation
are 10 to 100 times greater with pyridoxal phosphate than
with pyridoxal (Matsuo, 1957a; Metzler, 1957; Christensen,
1958).
Also, pyridoxal phosphate is 10 to 100 times more
reactive than pyridoxal toward semicarbazide (Cordes and
Jencks, 1962b), and toward a variety of carbohydrazides
(Wiegand, 1956).
Acid dissociation constants of the hydra-
zides were linearly related to specific reaction rates for
formation of pyridoxal phosphate hydrazones.
The greater
reactivity of pyridoxal phosphate than of pyridoxal toward
nucleophilic reagents, because of the existence of the lat­
ter compound largely in the form of the unreactive internal
hemiacetal, provides a further chemical advantage for
37
pyrldoxal phosphate as a coenzyme (Cordes and Jencks, 1962b).
Spectral methods for determining Imlne formation constants
Consider the reaction, A+P=^I, where A Is amino
acid, P Is pyrldoxal analog, I Is aldlmlne. The observed
constant, KQ = Ig/AgPg, where Ig, A^, P^ are the sums of
concentrations of all Ionic species of I, A, and P, respec­
tively, at equilibrium.
Early estimates of imlne formation constants were based
on molar absorbancy indices of solutions of imines, using
the equation: K = ( a - a )( a - a ), where a, a.., a
o
P
i
p
^ P
are apparent molar absorbancy indices of the solution, the
imlne, and the pyrldoxal analog, respectively, at the same
wavelength. Direct extrapolation of the spectral data was
used to obtain a^^ for imines from pyrldoxal phosphate
(Matsuo, 1957a; Christensen, 1958). Calculations of
and
formation constants by successive approximations were car­
ried out for imines from pyrldoxal (Metzler, 1957) and for
metal chelates of imines from pyrldoxal (Davis £t a]^., 1961).
The double reciprocal plot of Lucas et al. (1962) was
used to obtain absorbancles for imines of pyridine-4-aldehyde
(Brulee and French, 1964), and a slightly different equation
was used to calculate Kg:
- ( D - Dp )/A^( D^ - D ),
where D is absorbancy at 270 mp for imlne (i), total solu­
tion (D), and pyrldoxal analog in the solution (p), A^ is
the initial amino acid concentration in excess over that of
38
pyrldoxal analog.
"Affinity constants" (K
values) for imines
M
of pyridoxal phosphate and amino acids were measured by the
same graphical method at various pH values (King and Lucas,
1959; Lucas et al., 1962).
The equilibrium constant for the formation of aldimlne
from pyridoxal phosphate and cysteine was determined from
kinetic rate constants and from an equation relating absorbancles at a particular wavelength, initially (Do) and at equi­
librium (De), with different amino acid concentrations (Mackay,
1962).
In the former case. Ko = k^/k^, where k^, and k^ are
rate constants for imine formation and hydrolysis, respec­
tively. The direct spectrophotometric determination was based
upon the following relationships:
I = (D -D )/(a -a ) =
e
^ e o' i p
P»/K • (1 + 1/K A ). At two different concentrations of A,
°
° ®
D' ( 1 + 1 / K A" )
A' and A": I' /I" = (D' -D' ) =
0 ^
' o e \ And
e e
e
o
(1 + l/Vg)
solving for Ko = ^ o ^^'e"^'o)A'e - ^'o f®"e"^"o^/^"e . If
D'o (D"e-D\)
D"o (D'e-D'o)
A a n d A"^ are equal to the concentrations of A added (pseudo
first-order conditions), then Ko can be directly determined
from the spectrophotometric measurements.
A method of
successive approximations, less time consuming than the method
of Matsuo (1957a), was suggested if pseudo first-order con­
ditions did not prevail. In general, accuracy In this method
increases as Ko Increases (Mackay, 1962).
A similar equation was derived for measuring apparent
39
equilibrium constants for ketimine formation from pyridoxamine
and pyruvate (Banks et al., 1961).
Consider the above reaction
where A is now keto acid, P = pyridoxamine, I = ketimine. The
derived equation adapted to this situation follows: .
l/( D - D, ) = ( 1 + 1/K A ) . l/( a p - a P ), where D
0
1
0
p o i o
0
is absorbancy of P when A = 0,
is absorbancy of a solution
of P + I at some concentration A, P^ is initial concentration,
a^ = ( 2/P^ )(
)/K^A. Plotting l/(D^ - D^) vs. l/A
yielded a straight line, and
was found from the ratio,
intercept/slope. The precision of measuring
increased as
(DQ - D^) increased.
Molar absorbancy indices and apparent equilibrium con­
stants for imine formation from gamma-aminobutyric acid and
pyridoxal or pyridoxal phosphate were determined by the
graphical methods of Ketelaar _et
(1951, 1952) and of
Isenberg and Szent-Gyorgi (1958) (OHvo _et al., 1963).
These
methods are suitable for molecular complexes in which each
component participates with only one molecule, provided that
all reactants and reaction products obey Beer's law. As in
the previously described methods, all spectral data were
corrected for absorbancy of amino acid, but since the amino
acid concentration is high compared to imine concentration at
equilibrium, small decreases in amino acid concentration were
neglected in applying the correction.
Molar absorbancy
indices (ap) of pyridoxal and pyridoxal phosphate were deter­
mined at the desired wavelengths, temperature and pH, because
40
the contributions of various ionic forms of these compounds to
their total absorbancies change with the experimental con­
ditions.
Ketelaar's equation, 1/( a - a ) = l/K (a - a ) •
P
o
i
p
l/A + l/( a^ - a^ ) is derived in the Experimental, since it
was used for determining formation constants for imines from
pyridoxal raethochloride and valine, from pyridoxal phosphate
and leucine, and from deoxypyridoxal and leucine.
A plot of
l/( a - a^ ) vs. l/A yields a straight line, if pseudo firstorder conditions prevail.
The Ko value is the ratio of inter­
cept, l/( a. - a ), to slope, l/K ( a - a ).
^
P
°
i
p
An analogous method for determining equilibrium constants
for formation of complexes uses the following equation adapted
to this situation: l/( D - D. ) = l/( a - a ) • l/P +
Pi
i
p
o
K^/( a^ - a^ ) • l/B^ • 1/A^. A plot of l/(
) vs.
l/A^ gives a straight line with slope of K^/( a^ - a^ ) •
l/P^ and intercept of l/( a^ - a^ ) • 1/P^, and K
- slope/
intercept (Harbury and Foley, 1958).
The method of Isenberg and Szent-Gyorgi (1958) uses
differential spectral data and the equation:
( a^ - a^ ) P/(
l/A - K
o
- Dp ) - K^. The absorbancy of a solution
of imine is measured against a blank solution of pyridoxal
analog at the same concentration as in the imine solution
initially.
A plot of l/A vs. 1/(
) gives a straight
line with slope of K ( a. - a ), ordinate intercept of -K ,
o
1
p
o
and abscissa intercept of l/( a^ - a ) P.
4l
Calculations of K at different wavelengths yielded cono
cordant results, which was further evidence for correctness of
the assumption that observed spectral changes represented
imine formation and for proper application of Ketelaar's and
Isenberg's methods for measures of
and a^ (Olivo et al.,
1963). The fact that almost identical
values were obtained
by the two methods was evidence that
may be calculated from
observed spectral changes.
increased with tem­
Although
perature, a^ was independent of temperature (about 5 x 10^
for pyridoxal and gamma-aminobutyrate, 20-40°C).
Pyridoxal-N-methochloride imines
The spectral changes resulting from the addition of an
amino acid to pyridoxal methochloride have been interpreted
and compared to that resulting from formation of pyridoxal
imines (Johnston et al., 1963).
The imine equilibrium (or
formation) constant was directly calculated from the spectrophotometric data at pH 11, where the imine dipolar anion
(Pc-d, Figure 4; Table 2) is formed from an amino acid anion
'
and the pyridoxal methochloride dipolar ion. But at pH 6,
incomplete formation of the imine dipolar ion complicated the
direct determination of the equilibrium constant.
At pH 6,
decrease in the pyridoxal methochloride absorption band at
323 mp. corresponds with the loss of the hemiacetal dipolar
ion, and the simultaneous appearance of an absorption band at
4l9 m^, upon the addition of valine, corresponds with the hydro­
42
gen-bonded Imlne, from comparison with spectral studies of the
equilibria between pyridoxal and amino acids and their imines
(Metzler, 1957).
Atequilibrium in a solution with initial concentrations
of 10"^ M pyridoxal methochloride and 0.5 M valine the absorbancy at 4l9 m;a is about 0.7 as much as the absorbancy at 323
mjji. This was interpreted as a 30^ conversion of the pyridoxal
analog to the hydrogen-bonded valine imine.
The remaining
absorption band at 323 mp. was assumed to represent the 70^
unreacted pyridoxal methochloride.
Prom these assumptions the
extent of imine formation was estimated (Johnston _et
, 1963)
by solving for the fractions of total pyridoxal methochloride
in the free dipolar ion, imine dipolar ion and imine anion
forms by a successive-approximation method and from molar
absorbancy indices measured at wavelengths of maximum absorb­
ancy for each of these three forms.
The absorption maximum of 367
at pH 12 for the pyri­
doxal-valine imine dianion is shifted to 377 niji upon methylation of the pyridine ring nitrogen atom which thereby gains
a positive charge.
This is in the same direction but of a
different magnitude than the 15 #i shift of the absorption
band of the free pyridoxal anion upon protonation to form the
dipolar ion (Metzler and Snell, 1955). In general, the effect
of methylation of pyridoxal-amino acid imines on the absorption
maxima for other ionic forms is to shift these maxima about 5
myi to longer wavelengths (Johnston et al., 1963).
Although
43
the dipolar-Ionic valine imines of pyridoxal and of pyridoxal
methochloride do not differ appreciably in stability, the
hydrogen-bonded anionic imine of pyridoxal is approximately
100 times more stable than the non-hydrogen-bonded imine of
pyridoxal methochloride (Johnston et al., 1963).
Biological significance of imines of pyridoxal analogs
The pK for loss of a proton from the imine of pyridoxal
methochloride and valine was estimated to be 8.0, compared
with values of 6.2 for glutamic-aspartic transaminase and 7*3
for glutamic-alanine transaminase. These differences and
spectral differences between the enzymes and pyridoxal metho­
chloride imines suggest that if the enzyme-bound coenzymes
exist as a dipolar anion, their spectral behavior must be
profoundly affected by groups in the proteins (Johnston et al.,
1963).
Present knowledge of pyridoxal phosphate-enzyme spectra
has been recently reviewed (Velick and Vavra, 1962a; Jenkins,
1963; Guirard and Snell, 1964). Spectra of imines of pyri­
doxal analogs or of coenzyme analogs bound to apoenzymes may
help in the assignment of spectral absorption bands to the
various structural forms of pyridoxal phosphate enzymes.
Imines resulting from interaction of pyridoxal, deoxypyridoxal and pyridoxal phosphate with proteins, as well as
with amino acids and peptides, have been studied as to their
electronic absorption and infrared spectra, ionization and
formation constants (Table 3; Christensen, 1958; Dempsey and
44
Chrlstensen, 1962).
Table 3.
Spectra of imlnes of albumin and pyridoxal analogs
pH 7.5
Xmax
mji
a^ x 10"^
PL
317
415
6.80
1.45
DPL
339
426
2.22
2.81
PLP
332
413
3.71
0.94
Imine stability constants fell into three classes de­
pending upon the nature and specificity of binding groups.
The highest constants ranged from 10^ to 10^. Hydrogenbonded Imines (at 4l5 mp) were stabilized by phosphate
ester linkages to the protein, but were not specific for
phosphate esters.
Deoxypyridoxal, for example, reacted at
non-specific sites.
Pyridoxylamines (substituted imines) were
seen at 332 mja (Dempsey and Chrlstensen, 1962).
Affinity constants for the formation of imines between
pyridoxal phosphate and amino acids have been calculated from
changes in the absorption spectrum of pyridoxal phosphate
upon the addition of graded concentrations of amino acid and
compared to Michaelis constants of pyridoxal phosphate-
45
enzymes with their amino acid substrates (King and Lucas,
1959; Lucas et al., 1962; King, 1963).
A close correlation
was found for the complex formation of leucine with leucine
decarboxylase and with pyridoxal phosphate, which was inter­
preted to mean that the coenzyme may be. the binding site
which largely determines substrate affinity, and that the
apoenzyme does not greatly alter the stability of the
coenzyme-substrate link. The apparent pH optimum of 6.5
to T, which was found for the model imine, agreed with that
of the enzyme, and was considered due to à decrease in
substrate affinity at lower pH values (Lucas et al., 1962).
Also, inhibition of homoserine deaminase (a pyridoxal
phosphate enzyme) by amino acids paralleled the stability
constants of their imines with pyridoxal phosphate (Matsuo,
1957a).
The calculated rate of imine formation for aspartate in
glutamate-aspartate transaminase is at least 7.7 x lO^/M/min.
at pH 8.5, and may be compared with a value of 30/M/min. for
nonenzymic imine formation between aspartate and pyridoxal
phosphate (Cordes and Jencks, 1962b).
Only a slightly higher
value would be expected for the reaction of aspartate with a
pyridoxal phosphate imine. This strong catalytic influence
of the enzyme even on imine formation takes place perhaps
through intramolecular formation of cationic imine which is
more susceptible than an uncharged imine toward nucleophilic
amino groups in transaldimination.
46
In Figure 6, which compares (A) aldimine formation from
pyridoxal and an amino acid with (B) formation from a pre­
formed aldimine of pyridoxal (as in enzymic catalysis) by
transaldimination, the latter process is more efficient. It
is thus advantageous for pyridoxal phosphate-enzymes to have
the coenzyme bound as an imine, since imines of pyridoxal
and pyridoxal phosphate have been found to be at least thirty
times more reactive toward nucleophiles, such as semicarbazide
in semicarbazone formation, than the free aldehydes (Cordes
and Jencks, 1962b).
Therefore, the first covalent bond-
forming reaction with pyridoxal phosphate enzymes and amino
acids is transaldimination, since enzyme coenzyme linkage has
been shown to be an azomethine of the aldehyde group of pyri­
doxal phosphate and the epsilon amino group of a lysine
residue in the protein (Jencks and Cordes, 1963; Snell, 1962;
Guirard and Snell, 1964).
Synthetic imines (hydrazones, hydrazides and oximes) of
pyridoxal phosphate have been found, not only to behave as
coenzymes, for some pyridoxal phosphate enzymes, but in some
cases to be more active than pyridoxal phosphate itself, and
in others, to be more resistant to inhibition by strong
pyridoxal phosphate inhibitors, such as penicillamine (Gonnard,
1963; Gonnard et al., 1964; Makino et al., 1962, 1963; Ooi,
1964). Evidence for the biological roles of imines of pyri­
doxal phosphate (their roles in enzymic reactions) has been
reviewed (Braunstein, I96O; Snell, 1958, 1962; Guirard and
47
Snell, 1964).
However, only the role of these Imines in trans­
amination reactions with amino acids will be considered in this
review.
Function of Vitamin Bg in Transamination
"Transamination may be defined as the transfer of an
amino group from one molecule to another without the inter­
mediate formation of ammonia." (Guirard and Snell, 1964,
p. 138).
Metabolic significance of the transamination
reaction, including its central importance in intermediary
nitrogen metabolism, has recently been reviewed (Braunstein,
1957; Meister, 1962; Guirard and Snell, 1964). Its wide­
spread occurrence and role in biosynthesis of many amino
acids are among its important features.
Early nonenzymic models of transamination ignored the
coenzyme role of vitamin
(Herbst, 1944), and it was not
until 15 years after the discovery of the vitamin that its
coenzymic function was proposed (Snell, 1945b). This proposal
was based upon observed changes in specificity of lactic acid
bacteria for pyridoxal, pyridoxamine and pyridoxol, after
antoclaving pyridoxol (previously treated with potassium
permanganate) and glutamic acid (Snell, 1944; Snell and
Rannefeld, 1945), as well as with other amino acids; and
upon the demonstrated natural occurrence of pyridoxal and
pyridoxamine (Snell, 1945a; Rabinowitz and Snell, 1947).
The evidence at present for the coenzymic role of vitamin Bg
48
phosphate In transamination is quite substantial and has been
reviewed a number of times recently (Meister, 19^2; Snell,
1958; Snell and Jenkins, 1959; Braunstein, 196O; Sizer and
Jenkins, 1963; Guirard and Snell, 1964).
Mechanism of transamination
Early suggestions that imine intermediates were involved
in transamination (Braunstein and Kritzmann, 1937; Herbst,
1944), influenced the independent proposals for the classical
mechanisms for enzymic transamination (Braunstein and
Shemyakin, 1952, 1953), and for nonenzymic transamination
(Metzler et a]^., 1954a).
"Although the conclusions of these
two groups of investigators do not always coincide in details,
it seems Justified to label their propositions with the
general denomination of the theory of Braunstein, Metzler
and Snell." (Perault et al., 1961, p. 555).
The similarity of these enzymic and nonenzymic mech­
anisms, in which amino acid imines of pyridoxal phosphate or
of pyridoxal (or their metal chelates) are intermediates, is
in direct contrast to the different mechanisms proposed in an
early review of enzymic and nonenzymic transamination, neither
of which involved vitamin Bg (Herbst, 1944). The generally
accepted mechanism for enzymic transamination (sometimes
called the binary mechanism or pyridoxal phosphate-pyridoxamine phosphate "shuttle mechanism") takes into account the
established nature of coenzyme binding and resolution, recent
49
kinetic studies of transaminases, and results from investi­
gations on nonenzymic model systems, all of which have been
adequately reviewed recently (Lis et al., I96O; Jenkins and
Sizer, 196O; Velick and Vavra, 1962a, b; Snell _et
, 1963;
Hammes and Pasella, 1963b; Guirard and Snell, 1964). The
descriptive details of this mechanism have been clearly pres­
ented recently (Snell and Jenkins, 1959; Snell, 1962; Jenkins,
1963; Jencks, 1963; Pullman, 1963; Guirard and Snell, 1964)
and summarized in Figure 7.
Initial activation of the amino acid is postulated to
occur through formation of an aldimine between the amino acid
and pyridoxal (or a pyridoxal-containing enzyme). Labilization
of the bonds to the alpha-carbon atom of the amino acid leads
to the several reactions, including transamination, that are
catalyzed by the versatile vitamin, or more efficiently by
enzymes that contain it (Guirard and Snell, 1964). However,
the only differences in the enzymic and nonenzymic mechanisms
are:l) imine formation and hydrolysis in the nonenzymic mech­
anism is replaced by the more efficient transamination process
in the enzymic mechanism, and 2) the nature of the catalytic
groups effecting the aldimine-ketimine tautomerization and the
stereospecificity in the enzymic reaction on the substrate are
difficult to duplicate in a model system.
After transaldimination the liberated epsilon-amino group
of lysine may act as a general acid-base catalyst for the
electron shifts initiated by the pyridoxal coenzyme (Snell,
Figure 7. Classical binary or shuttle mechanism of pyridoxal phosphate function
in transamination (Snell, 1962; Guirard and Snell, 1964)
Steps: 1.
Approach of non-hydrogen bonded aldimine form of enzyme to amino
acid
2,3. Transaldimination
4,5.
6.
Tautomerization (rate-determining of aldimine to ketimine)
Hydrolysis
Other suggested absorption bands for enzyme-substrate intermediates
for glutamic-aspartic transaminase (Sizer and Jenkins, 1963; Jenkins,
1963; Hammes and Fasella, 1963a; Guirard and Snell, 1964):
Non-H-bonded aldimine (upper left) 362 mp.
X (upper right)
330 mja
H-bonded aldimine (lower right)
425 - 35 mp
Y (lower center)
492 mjm
Ketimine
325 mp
Pyridoxamine form (lower left)
332 mjx
• H.9 XOf®
IT
ÇHNHC
•HN-C=0 0
I
II
e hSD-P-O
Ô
—
—
©
W°v®
±
R J.CO2®
©
z
;
o
QC
CL
•B:H
NON H-BONDED FiLDlMlNE ^62
VJ1
M
KETO FCID^
R^COg®
0
NH-j
©^::
PYRIDOXFIMINE
FORM (332myu)
e
R. .cog;
—NHg
©
—NHg
,N
4
—BH
KETIMINE
H-BONED FiLDlMlNE
(425-aOny)
52
1962).
This suggestion of the role of the epsilon-amino
group of lysine was meant to be symbolic of the way in which
one or more proximal nucleophilic groups on the protein may
facilitate the labilization and transfer of protons.
"At our
present level of knowledge, such schemes ignore the unknown
factors responsible for substrate and reaction specificity
of individual pyridoxal phosphate-proteins, and undoubtedly
oversimplify as well the catalytic events occurring on the
coenzyme." (Guirard and Snell, 1964, p. 158).
The prototropic shift (or removal of the alpha-hydrogen),
converting the aldimine to the ketimine form, has been shown
to be the rate-determining step in enzymic, as well as in
nonenzymic transamination systems, as predicted (Metzler,
1957) and recently substantiated (Velick and Vavra, 1962a,b;
Hammes and Pasella, 1963a,b; Jenkins, 1963; Banks et al.,
1961, 1963; Blake e^ aj., 1963; Bruice and Topping, 1963a,d;
Junk and Svec, 1964; Vernon, 1964).
General base-catalysis
facilitates the removal of the proton from the alpha-carbon
atom of the amino acid residue in the aldimine form, and
general acid catalysis may enhance protonation of the formyl
carbon atom, as well as the pyridine nitrogen atom. Hydro­
lysis of this resulting ketimine leads to formation of an
alpha-keto acid corresponding to the amino acid substrate.
Or, another keto acid may exchange with this coenzyme-bound
keto acid in a process of transketimination, followed by the
prototropic shift, and resulting in a new amino acid after
53
hydrolysis of this second aldimine, thus completing a trans­
amination between two amino and two keto acids in a con­
secutive manner by way of binary complexes with the coenzyme
on the enzyme surface.
A positively-charged nitrogen atom of the pyridine ring
acts as an "electron sink" to facilitate electron withdrawal
from the alpha-carbon atom of the amino acid in the aldimine
(Cordes and Jencks, 1963).
The original Metzler-Snell-
Braunstein hypothesis proposed that the inductive effect of
this electronegative group weakens the bonds around the alphacarbon atom in the aldimine through a reduction of electron
I
density in these bonds. Rupture of one of these bonds with
formation of a conjugated system of double bonds extending
from the alpha-carbon atom to the electronegative group
(pyridine ring nitrogen atom) gives an intermediate in which
the "extra" electron pair can be localized in any of a number
of possible ways (Metzler et a^., 1954a; Braunstein, 196O;
Guirard and Snell, 1964).
However, a modification in the interpretation of the
driving force of this mechanism has been made on the basis of
molecular orbital (LCAO) approximations (Perault et al., 196I;
Pullman, I963).
These approximations indicate that the
electron density about the alpha-carbon atom in the aldimine
is not reduced, but is actually increased. Furthermore, this
carbon atom is not a part of, but outside of, or next to the
conjugated system, since it is a saturated carbon atom in the
54
aldimine. The driving force for prototroplc shift in the
interconversion of alpha-amino acid aldimines of pyridoxal is
then the stabilization of an intermediate carbanion owing to
a gain in resonance energy by creating a particularly favor­
able conjugated system. That is, resonance stabilization of
an intermediate carbanion drives the tautomerlzatlon of imlnes
in transamination.
Since appreciable quantities of ketoenamine exist in
solutions of pyridoxal Imlnes, the participation of the
protonated imine nitrogen atom as a strong electrophillc
center has been suggested as an additional driving force in
the mechanism of pyridoxal-catalyzed reactions (Brulce and
Topping, 1963d; Jencks, 1963). The most reactive species of
pyridoxal phosphate Imlnes towards transamination is pro­
tonated at the imine nitrogen atom (Cordes and Jencks, 1962b;
Jencks and Cordes, 1963).
This protonated nitrogen atom,
adjacent to the two carbon atoms Involved in proton exchange
during transamination, may facilitate the rate-determining
loss of the proton from the adjacent saturated carbon atom to
I
form the transition state. The electronic Influence of a
proton or metal ion attached to this imine nitrogen atom
facilitates formation of such a transition state with a
positive imine nitrogen atom (Martell, I963).
The Inductive
effect of the pyridine ring nitrogen atom assists in the
formation of the transition state which is favored by the
positively-charged nitrogen atom in the enamine. This inter­
55
pretation of the mechanism also applies to the metal ioncatalyzed transamination reactions (Metzler and Snell,
1952b) and to the transamination reactions in ethanol (Matsuo,
1957b), since significant amounts of the ketoénamine form are
expected in both model systems.
Kinetics of the transaminase reaction has supported a
Ping Pong Bi Bi mechanism (i.e., Binary complexes between
amino acid or keto acid and coenzyme are formed and exchanges
of the amino and keto groups take place between coenzyme and
substrate.), and has directly contradicted the suggestions
(Evangelopoulous and Sizer, 19^3) that the enzyme has only a
pyridoxal phosphate coenzyme form and that the enzyme forms
a ternary complex with both amino and keto acids prior to
exchange of amino and keto groups (Benson and Cleland, 1964).
Although the modified classical mechanism of coenzyme action
in transamination is unquestionably insufficient to explain
the way in which the protein so greatly enhances the catalytic
effectiveness of the coenzyme, this theory has most of the
evidence in its favor, has correlated a large body of data,
and has suggested many further experiments, thus proving its
usefulness.
Nonenzymic Transamination in Model Systems
The purpose of a model system is to simulate the real
system found in nature, at least in the particular aspects to
be studied, and to provide a simpler subject for convenience
56
of measurement.
Model transamination systems have consisted
of either pyridoxal phosphate, pyridoxal or one of its
structural analogs reacting with an amino acid, peptide or
other amino acid derivative, or pyridoxamine or its phosphate
coenzyme reacting with a keto acid, under suitable conditions,
which may include various additional catalysts.
These non-
enzymic systems have provided simple and suitable media for
detailed studies of the mechanism of the enzymic trans­
amination, of the role of the coenzyme and of its structural
requirements in the reaction. However, these chemical models
are related to the enzymic reaction only if it is known how
readily pyridoxal phosphate or pyridoxal combines with amino
acids, how chelating metal ions Influence the reaction, and
what relation, if any, exists between the affinity of pyri­
doxal phosphate and an amino acid under physiological con­
ditions and the affinity of pyridoxal phosphate enzymes for
amino acids (Lucas
, 1962). For recent reviews on non-
enzymic reactions of pyridoxal with amino acids and their
significance see the following: Snell, 1963; Snell, 1962;
Westheimer, I96O; Brulce and Topping, 1963a; Cennamo, I963;
Guirard and Snell, 1964.
When pyridoxal and pyridoxamine became available in pure
form, Snell first demonstrated the nonenzymlc transamination
between these vitamin Bg compounds and glutamate and alphaketoglutarate at physiological pH values in dilute aqueous
solutions at autoclave temperatures (Snell, 1945b). The
57
reaction occurred with other amino and keto acids, with pyrldoxal and pyrldoxamlne, acting as an amino group acceptor and
an amino group donor, respectively, In the reversible trans­
formation of an amino acid to a keto acid.
The rate of the
reaction was Increased by as much as 20-fold by the addition
of appropriate metal Ions, of which the most effective were
Cu (II), A1 (III), and Fe (ill) (Metzler and Snell, 1952b;
Gregerman and Chrlstensen, 1956; Longenecker and Snell,
1957), and by as much as 100-fold by Cu (II) at pH 10, room
temperature (Banks et al., 1961).
Nonenzymic, metal Ion-
catalyzed transamination between peptides and pyrldoxal has
also been demonstrated (Cennamo, 1954, 1958). The activity
of metal Ions In catalyzing these reactions Increased In
approximately the order of Increasing stability of the
pyrldoxamlne chelates with the corresponding metal Ions
(Longenecker and Snell, 1957; Gustafson and Martell, 1957).
The roles of metal Ions In Imlne chelate formation and In
the catalysis of nonenzymic transamination have been described
(Metzler _et a]^., 1954; Snell, 1958, 1962; Brulce and Topping,
1963a; Gulrard and Snell, 1964). Similarity between reactions
catalyzed by pyrldoxal phosphate enzymes and those catalyzed
by pyrldoxal and metal salts has been stressed (Metzler, e^ al,
1954a).
Metal Ions have been considered as playing Ineffi­
ciently In nonenzymic systems the functional role played by
acidic and basic groups In the protein In enzymlc systems
58
(Metzler et al., 1954a; Snell, 1958, 1962; Snell and Jenkins,
1959; Perault et al., 196I; Guirard and Snell, 1964).
Evidence that such nonenzymio transamination reactions
proceed through inline intermediates has included spectrophotometric (Eichorn and Dawes, 195^; Metzler, 1957, Matsuo,
1957a; Pasella e;b
, 1957), electrophoretic or chromato­
graphic detection of reactions, aldimine and ketimine inter­
mediates, and products (Pasella _et ^., 1957; Matsuo, 1957b;
Banks et al., 196I; Olivo et^., 1963; Cennamo, 1963).
The
nonenzymic transamination of pyridoxal phosphate by an excess
of amino acid to pyridoxamine phosphate at physiological
temperature (37°G), in the presence of metal ions, was
observed chromatographically and spectrophotometrically.
At pH 5 there was good reaction because chelated imines were
unstable, but at pH 7 there was poor reaction because chelated
imines were more stable (Cattaneb et al., i960).
Although at least a portion of the nonenzymic reaction
proceeds through the metal chelated imines when metal ions
are present, as shown by several semi-quantitative kinetic
studies (Pasella et al., 1957; Cattaneb et al., i960), the
reaction proceeds slowly in water (Metzler, 1957; Metzler and
Snell, 1952b; Banks et al., 196I; Fleck and Alberty, 1962)
and rapidly in ethanol (Matsuo, 1957b; Bruice and Topping,
1963a) and in highly alkaline medium (Gustafson and Martell,
1957) without the addition of metal ions.
In fact, metal
ions inhibit reaction in ethanol (Matsuo, 1957b) and do not
59
seem to affect the reaction rate in alkaline solution in the
presence of high buffer concentration, such as 1.8 M imidazole
(Bruice and Topping, 1962, 1963a).
Another metal ion-independent nonenzymic transamination
has recently been established for pyridoxal and leucine ethyl
ester (Cennamo, 1964). This reaction proceeded at a rate
comparable to the alum-catalyzed reaction between pyridoxal
and leucine.
Although the transamination of amino acid esters
with pyridoxal proceeded to greater than 90^ decrease in pyri­
doxal after 40 minutes at 100°C, pH 5, the reverse reaction
was also uncatalyzed with metal ions; nor was it enhanced by
keto acid esters (Cennamo, 1961, 1964). Reactions of pyridoxal analogs with leucine ethyl ester at 25 C have indicated
that deoxypyridoxal and pyridoxal phosphate are much better
amino group acceptors than pyridoxal and that favorable con­
ditions are in slightly acidic solution (Johnson, 1964).
Furthermore, less imine formation was observed for the
amino acid esters than for free amino acids and metal ions,
which suggests that the former imines are more reactive (less
stable) than the latter (Cennamo, 1962).
Although metal
chelation allows formation of high concentration of imines,
most of these chelates may be poor reactive intermediates
(Cennamo, 1963).
These reactions of pyridoxal with amino acid
esters have provided experimental evidence for the concept
that one role of the metal ion in imine chelates is to increase
the electronegativity of the carboxyl group by suppressing its
60
Ionization (Snell, 1958), which is what esterification does.
An essential feature of enzj/raic transamination is the binding
of the carhoxyl group of amino acids to the apoenzyme
(Jenkins^
, 1959).
On the other hand, lack of an
Inductive effect of the carboxyl group in gamma-aminobutyric
acid caused Inadequate activation of protons bound to the
gamma-carbon atom of this amino acid in an imine with pyridoxal, and resulted in a very slow transamination reaction
(Olivo et^., 1963).
Imidazole catalyzed a transamination between phenylglycine
and pyridoxal in aqueous solution that was even more rapid than
the metal-ion-catalyz^d reaction (at least at pH 8.6)(Bruice
and Topping, 1962, 1963a,b,d). This finding is of potential
significance because of the presence of imidazolyl and other
polar groups of proteins, perhaps acting as catalysts in
reactions of the imines of substrates and pyridoxal phosphate
enzymes, and because of the lack of metal ions in purified
transaminases and the lack of their effect in enzymic trans­
amination (Pasella _et a]^., 1962; Snell and Jenkins, 1959j
Jenkins et al., 1959).
This model for the pyridoxal-requiring transamination
reaction has other features of enzymological interest: l)
The ability of imidazole to form a catalyst-substrate complex
resulted in a virtual specificity for imidazole at low reactant
concentrations compared with other general bases Investigated,
such as carbonate and morpholine. The formation constants for
61
the complexes of Imidazole with phenylglycine were found to be
constant over the pH range 7 to 10, although the over-all rate
of reaction with pyridoxal increased with an increase in pH
(Bruice and Topping, 1963b). The pK^^ of 7.78 for the maximum
velocity (V^) as a function of pH approaches the pK^^ of 8.3
for pyridoxal (Metzler, 1957). 2) The reaction operated in
aqueous media at physiological pH values (j to 10} and
temperatures (30°).
3) The transamination reaction in this model system was
equally effective if preceded by a transamination reaction,
the transamination of phenylglycine and pyridoxal proceeded
at the same rate if phenylglycine was added to the imine of
pyridoxal and morpholine (Bruice and Topping, 1963c).
The
catalysis of pyridoxal reactions by morpholine, which must
involve the intermediate with a protonated imine nitrogen
atom, was highly effective in the case of pyridoxal phosphate
but less effective and levelled off in concentrated morpholine
solutions in the case of pyridoxal.
This levelling off, and
the spectral changes observed upon the addition of morpholine
to pyridoxal, suggested that an unreactive cyclic aminoacetal
was formed (Cordes and Jencks, 1962b).
4) Catalysis was by weakly basic imidazole and weakly
acidic imidazolium ion, in which the proton of the alpha-carbon
atom of the amino acid residue in the aldimine was abstracted
by the general base and a proton was donated to the azomethine
bond by the general acid species of imidazole, to complete the
62
concerted general acid, general base catalysis of the Intracomplex prototroplc shift (Brulce and Topping, 1963d).
In the
enzymlc reaction the abstraction of the proton must be carried
out by the weakly basic groups available to the protein at pH
values near neutrality, the most effective of which would be
the imldazolyl group of hlstldlne.
If it was assumed that
dissociation constants of all the complexes with imidazole and
imidazolium ion are similar, that formation constants for
aldimines are in the usual range of 1 to 100 (Metzler, 1957),
and that the acid dissociation constants of the imine inter­
mediates are similar, then it was calculated that the rate
of prototroplc shift leading to conversion of aldlmine to
ketlmlne was only about 100 to 1000 times slower than the
corresponding step for glutamate-aspartate transaminase
(Brulce and Topping, 1963c).
The rate constant for this pro­
totroplc shift was calculated to be 4 to 200/min. (Brulce and
Topping, 1963d).
Other comparisons of enzymlc and model transamination reactions
The imine formation constant for pyridoxal and valine is
about 10^ times less than that for transaminase and an amino
acid and the rate constants for the reaction of these imines
differ by a factor of 10® in favor of the enzyme. This
indicates that the enzymes catalyzes both the transamination
step and the tautomerization step (Cordes and Jencks, 1962b;
Brulce and Topping, 1963d; Hammes and Pasella, 1963a).
Even
63
though the gross aspects of model systems are similar to the
enzymic reaction, the details are obviously quite different.
The role of the protein molecule is unknown; there may be
static and dynamic conformational effects which bring the
substrate into critical configuration for polarizing appropri­
ate chemical bonds.
If acid-base catalysis is involved the
reaction rate is limited by the rate of dissociation of acidic
or basic groups.
If the pK^^ for such a critical group is
between 6 and T, the maximum rate of acid-base catalysis is
about lO^/sec. assuming a diffusion-controlled rate of pro­
tonation. And the interconversion of imines is slow enough
for acid-base catalysis to be involved.
Dynamic conformational changes may cause physical strain­
ing of bonds and the rate of such a change could be kinetically
important. It is known that the basic form of the enzyme is
necessary for the catalytic interconversion of intermediates,
but not for imine formation. This catalytic interconversion
is of the order of 10^/^/seo. and the rate of aldimine to
ketimine transformation was calculated to be between 10 and
100/sec. (Hammes and Fasella, 1963a).
The reaction of free aspartate and pyridoxal phosphate
imines is too slow to account for the over-all rate of enzymic
reaction, and consequently, the transamination between aspar­
tate and enzyme-bound pyridoxal phosphate must be catalyzed by
glutamic-aspartic transaminase, possibly by conversion of the
pyridoxal phosphate-enzyme imine to the much more reactive.
64
protonated species (Cordes and Jencks, 1962b). However, this
enzyme reacts with alanine at pH 8 with a first-order rate
constant of 0.20/min., a value similar to that for the nonenzymic reaction of pyridoxal phosphate with aspartate (Jenkins,
1961).
Also, glutamic-aspartic transaminase lost its ability
to catalyze the transamination reaction to the extent of its
resolution to coenzyme and apoenzyrae. The apotransaminase
formed a holoenzyme analog with pyridoxal and catalyzed the
reversible transamination of pyridoxamine and alpha-ketoglutorate or oxaloacetate, the half-reactions of the over-all
enzymic transamination. But the efficiency of catalysis by
this holoenzyme analog was only about 0.1^ of the rate of the
corresponding reaction catalyzed by the natural holoenzyme,
a finding which emphasizes the importance of firm and
sterically-determined coenzyme-apoenzyme binding.
This
markedly enhanced efficiency of the enzymic process was made
possible by eliminating the several additional reactions
involved in binding and releasing free pyridoxal or pyri­
doxamine after each interaction with substrate (Wada and Snell,
1962).
In summary, the catalytic effects of the enzyme on the
transamination reaction has recently been estimated as 10^
times faster than the tautomeric interconversion of aldimine
and ketimine in free solution, at zero buffer concentration
and in the absence of other intermolecular catalysts (Vernon,
1964).
65
Structural requirements In pyrldoxal for
nonenzymlc transamination
In addition to providing simple systems permitting the
study of aldlmlne and ketlmlne Intermediates In the trans­
amination reaction, the nonenzymlc reaction also provides a
convenient experimental system In which the structural re­
quirements for participation In amino group exchange reactions
and hence for catalysis of the over-all transamination reaction
can be studied. By studying many aldehydes as amino group
acceptors In such model systems the Importance of each of
the functional groups of pyrldoxal In the over-all reaction
was assessed (Ikawa and Snell, 1954a; Metzler £t
, 1954a,b).
The essential functional group requirements are the following:
l) a formyl group, since it is converted during trans­
amination to the aminomethyl group of pyrldoxamine and since
compounds lacking a carbonyl group do not accept amino groups
from amino acids. Its function is to form Imlnes with the
amino acids.
2) A heterocyclic nitrogen atom provides a strongly
electrophllic grouping which aids the catalysis of reactions
of the imlnes.
The nitrogen atom needs to be in an ortho- or
para-position to the formyl group on the pyridine ring for
maximum effectiveness.
3) A phenolic group must be unsubstltuted, since
substitution, or lack of a phenolic group resulted In greatly
decreased reaction rates. The free phenolic group should be
66
in an ortho-position to the formyl group, since 2- or 4hydroxypyridine possessed the minimum structural requirements
for catalysis of nonenzymic transamination and similar
reactions by substituted pyridines (Metzler et al., 1954a).
A possible role of the phenolic group is stabilization of thé
imine intermediate through hydrogen-bonding or metal ionchelation, thus tending to maintain the system of conjugated
bonds in a planar configuration and favoring the requisite
electron shifts.
The hydroxymethyl group at the ^-position of pyridoxal
is not necessary for nonenzymic reactions; in fact it causes
a decrease in reaction rate, as compared to analogs lacking
the hydroxyl group (as in deoxypyridoxal) or having it
substituted (as in pyridoxal phosphate), since it permits
internal hemiacetal formation with the adjacent formyl group,
thus reducing the concentration of the reactive aldehyde in
solution.
The hydroxymethyl group is essential, however, for
formation of the coenzyme, and therefore, for enzymic reactions.
The phosphate ester group not only prevents hemiacetal for­
mation and maintains a high level of free aldehyde, but also
has an important role of contributing to the binding of coenzyme
to apoenzyme (Banks et al., 1963; Sizer and Jenkins, 1963; Vada
and Snell, 1962).
The methyl group in position 2 appears to be non-essential,
however, through its inductive effects, it may alter the re­
action rate (Marvel and Tarkoy, 1957). Also, omega-methyl-
67
pyridoxal-5-phosphate (PLP with an ethyl instead of a methyl
group at position 2) provided coenzymic activity for some,
but not for other apoenzymes, indicating wide differences in
the nature of binding sites of individual apoenzymes (Olivard
and Snell, 1955a,b).
A correlation of inhibitory potencies
of various coenzyme analogs for different PLP-enzymes has
revealed that every substituent grouping of PLP affects the
firmness of binding of coenzyme to apoenzyme (Snell, 1958),
Finally, a recent suggestion has been made for the role of
the electrophilic site around position 6 and the nitrogen
atom of the heterocyclic ring (Makino et al., 1963).
Analytical methods
Quantitative analytical methods used in following nonenzymic transamination reactions between pyridoxal and amino
acids have Included spectrophotometric measurements of elec­
tronic absorption bands associated with reactants, imine
intermediates, and products: measurements of optical rotation
changes resulting from racemization of the amino acid, and
chemical methods for reacting with remaining reactant or
forming product, followed by a colorimetric or spectrophoto­
metric determination.
Qualitative methods for detecting the
presence of products, intermediates, and reactants have
included paper chromatography and electrophoresis.
Spectral changes have been followed during model trans­
amination reactions with increasing absorbancy at 246 mp.
68
corresponding to ketimine formation, and with increasing
absorbancy at 395 mp.) corresponding to pyrldoxal loss (Bruice
and Topping, I962, 1963a); with decreasing absorbancies at
440, 4lO, 380, 320 m)i, corresponding to loss of pyrldoxal
or its aldimlne intermediate (Fleck and Alberty, 1962); with
decreasing absorbancies at 4l4 and 200 m)i, corresponding to
aldimine-metal chelates of pyrldoxal phosphate, and with
increasing absorbancy at 355 mji, corresponding to ketimlnemetal chelates of pyridoxamine phosphate (Cattaneo, et al.,
i960); and with changing absorbancy at 370 nya, corresponding
to aldimine-metal chelate of pyrldoxal, and with decreasing
absorbancy at 317 m)i, corresponding to pyrldoxal hemiacetal
(Cennamo, I963).
Optical rotation changes were measured at
546 m)i by the symmetrical-angle method (Fleck and Alberty,
1962).
Chemical methods have included analysis of pyrldoxal by
forming an ethanolimlne with maximum absorbancy at 365 np
(Metzler and Snell, 1952b; Cennamo, 1961, 1962, I963, 1964;
Banks _et
, 196I; Gregerman and Chrlstensen, 1956) and by
forming a red condensation product from two moles of pyrldoxal
and one mole of acetone in alkaline solution (Slegel and
Blake, 1962; Blake
, 1963).
Other methods which could
be used for analysis of pyrldoxal analogs in nonenzymlc
systems include condensation with cyanide (pyrldoxal phosphate
was determined as a cyanohydrin derivative at its maximum
absorbancy of 385 m)i by Scardi and Bonavlta, 1957), or with
69
phenylhydrazlne (phenylhydrazones of pyrldoxal and pyrldoxal
phosphate had identical maximum absorbancy Indices at 4lO mp.,
as measured by Wada and Snell, 1962).
Analysis of pyrldoxa-
mlne has been made spectrophotometrlcally after separation
of the 2f 4-dlnltrophenylhydrazones of pyrldoxal and keto
acid In a complex procedure (Metzler and Snell, 1952b; Banks
et al., 1961).
Analysis of keto acid has been carried out by methods
using a nonspecific chemical reaction with dlnltrophenylhydrazlne, followed by spectrophotometrlc measurement after
separation of the pyrldoxal hydrazone derivative (Metzler and
Snell, 1952b; Gregerman and Chrlstensen, 1956; Banks et al.,
1961).
Recent reviews have evaluated the relative advantages
and disadvantages of analytical methods for keto acids (Nelsh,
1957; Splkner and Towne, 1962; Robins et al., 1956).
Chemical reagents superior to the outmoded phenylhydrazone
methods because of their much higher degree of specificity
for alpha-keto acids Include 3-quinolylhydrazine (Robins^
al., 1956) and ortho-phenylenedlamlne (Bruice and Topping,
1963a; Splkner and Towne, 1962).
Qualitative identifications of intermediates and products
have been made by electrophoresis (Fasella £t
, 1957) and
by paper and column chromatography (Metzler and Snell, 1952a;
Metzler ejb aJ., 1954b; Fasella _et
, 1957; Kalyankar and
Snell, 1957; Catteneo et al., 196O; Cennamo, 1954, 1962, 1963,
1964). Detection was made by fluorescence techniques and by
70
chemical reagents, such as dlnitrophenylhydrazlne yielding
yellow products with keto acids, nlnhydrln yielding an orange
product with pyrldoxamlne and Indigo or violet products with
amino acids, and ammonia or ethanolamlne yielding a yellow
product with pyrldoxal. Recently, thin-layer chromatography
and spectrophotometry of alpha-keto acid hydrazones have been
studies In detail (Dancls _et £l., 1963).
Kinetic methods
Earliest rate measurements of nonenzymlc transamination
reactions were merely comparisons of changes In concen­
trations of pyrldoxal, pyrldoxamlne, or keto acid In a certain
period of time under different conditions (e. g., Metzler and
Snell, 1952b). Although many of these qualitative rate
studies were made, quantitative rate constants were not
reported until 1961 by Banks _et al. Their derived rate
expression Included the imine formation constant: l/v =
l/kA^ • (l/K^P^+ 1) + l/kP^, where v is the initial re­
action velocity, k is the pseudo first-order rate constant
for conversion of the ketimlne to pyrldoxal and alanine, in
the reverse reaction between pyrldoxamlne (PQ) and an excess
amount of pyruvate (A^). The formation constant for ketimlne
was found from a plot of l/v vs. l/A
= l/(slope/lntercept-
Po).
However, the complexities of their system and the
slowness of the reactions prevented a determination of order,
etc., and only allowed a determination of initial rates.
71
because the reactions could be followed to a maximum of only
6% completion, according to Bruice and Topping (1963a).
These reactions were not simple since both rapid and slow
spectral changes occurred on mixing the reactants (Matsuo,
1957; Christensen, 1958; Blake et al., 1963).
These changes
were functions of pH, temperature, concentrations of reactants,
buffer or metal ion catalysts, and indicated a number of
intermediates, as well as different species from reactants
(e. g., hemiacetal, hydrated and unhydrated aldehyde forms
in various ionic states) (Banks _et ^., 196I; Blake et al.,
1963).
At relatively low amino acid concentrations the
reaction rate obeyed pseudo first-order kinetics, but at high
amino acid concentrations, zero-order kinetics were followed,
as in enzymic catalysis (Banks et al., 1961). By increasing
the temperature 75°, to lOO^C, it was found that the reaction
went practically to completion and was first-order irl pyridoxal and in alanine (Blake et al., 1963). (Kinetics of
buffer and metal ion catalyses effects of pH, concentrations
of reactants, and ionic strength; stoichiometry, and com­
parisons of rate constants under various conditions for these
model systems are discussed later: see Results and Discussion).
Guggenheim plots were used to obtain pseudo first-order
rate constants for the nonenzymic transaminations of pyridoxal
with alanine (Fleck and Alberty, 1962) and with phenylglycine
(Bruice and Topping, 1963a). (See Experimental for further
discussion of the Guggenheim method, its advantages and
72
suitability for these reactions.) In the former reaction
three spectral relaxation times were found. ïTiese reaction
steps supposedly corresponded to formation of carbiholamine,
aldimine and ketlmine from alanine and pyridoxal. When the
initial alanine concentration was much greater than the
initial pyridoxal concentration, the absorbancy, A , due to
V
pyridoxal. Intermediates and products, is given as a function
of time by Ax. = B + Be
B e'^2^ + B
^j^gre m ,
o
1
2
3
1
mg, and m^ are pseudo first-order rate constants in the order
of decreasing magnitude.
Kinetics for the latter imidazole-catalyzed reaction
(Bruice and Topping, 1963a) followed a rate equation which
depended upon the square of the imidazole concentration.
This form of the equation is of the Michaelis-Menten type for
enzyme kinetics and represents saturation of an intermediate
with two molecules of an imidazole species, or alternatively
like saturation of a rate-determining step:
^obs ~ 0'0095 (lm)^/0.20 + (im)^, where k^^^ is the observed
or pseudo first-order rate constant for ketimine formation and
(Im) is the total imidazole buffer concentration. The equation
derived for the observed first-order rate constant for
attainment of equilibrium was the following: k ,
= k:K
1 1
(2c/2.303) = l/t log (.2_I_H) + B, where k is the first-order
Ob s
rate constant for the rate-determining, imidazole-intracomplex, aldimine-ketimine tautomerization,
and
are
equilibrium constants for formation of aldimine and its
imidazole complex.
73
- ab) + l/(K K )' 1/2, a Is the
1 2'
4
pyridoxal concentration and b is the amino acid concentration
c = a^+
+ 2(^ * ^
KlKg
at t^ and at any time t, ketimine concentration is x, pyri­
doxal concentration = a - x, amino acid concentration = b - x,
since aldimine was at a low steady state concentration, was
rapidly established and couldn't be measured spectrophotometrically; u = x -
^
g = -ic In (C - ^
c -f u - X
2
(c - u) = o at tg.
These two equations for k^^g were combined to provide
the following expression for the second-order rate constant,
kg, for initial equal reactant concentrations (a = b = 10"^ M):
kg =
(0.20 +
= 1.15/ct log ( f - ^ )
B = 0.0218 K^Kg ( I m f /
(0.0004 K^K + l)^/^.
At higher imidazole
buffer concentrations than about 1.8 M the rate constant
merely depended upon K, and K_: k = 0.0218 K K /(0.0004
,
J^
c
JL 2
K^Kg + 1)^/^.
74
EXPERIMENTAL
Materials
Pyridoxal analogs
Pyrldoxal-N-methochlorlde (PLM-Flgure 1) was synthe­
sized and found to be of good quality by elemental analysis
(Johnston et al., 1963). Deoxypyridoxal (DPL-Pigure l) was
synthesized by an improved method and purified by recrystallization from water by Dr. Isao Tomita in this laboratory.
Deoxypyridoxal was also purified by sublimation at about
40°C, atmospheric presure ( a^ = 6.45 x 10^ at 294.5
0.1 N HClj lit. value; a^ = 6.32 x 10^, pH 1; Metzler and
Snell, 1955; Heyl et
, 1953). Dr. Tomita also supplied
samples of 5-"carboxypyridoxal" or "5-pyridoxalylic acid"
(CPL-Figure l) and "alpha^-pyridoxalylacetic acid" (APIFigure 1; Tomita and Metzler, 1964). These pyridoxal
analogs were synthesized according to the methods presently
unpublished and outlined in Figures 35, 36 (Results and
Discussion).
Pyridoxal phosphate monohydrate (980 by assay) was
obtained from Sigma Chemical Co. (observed a
295 m|i, pH 1; lit. value: 6.70 x 10^; a
= 7.18 x 10^ at
m
= 5.27 x 10^ at 388
mp., pH 7; lit. value: 4.90 x 10^; a^ = 2.47 x 10^ at 330 mp.,
pH 7; lit. values: 2.45 to 2.50 x lO^; refer to Table 1,
Review of Pertinent Literature for sources. These spectral
75
data Indicated that this source of pyridoxal phosphate was of
high quality.
Pyridoxal hydrochloride (PL) was also obtained
from Sigma Chemical Co.
In general, pyridoxal analogs were
stored in a freezer at -20°C and warmed to room temperature
in a desiccator.
Amino acids, keto acids, buffers, and other chemicals
Amino acids, buffers and other chemicals were obtained
from commercial sources.
Leucine was recrystallized from
ethanol and water after treatment with activated charcoal
prior to use in measurements of imine formation constants,
in order to remove impurities which have a small absorbancy
below 350 m}i.
However, no noticeable effect of these
impurities on the over-all rate constants for nonenzymic
transamination reactions with pyridoxal analogs was observed.
Especially when a high concentration of leucine was desired,
such as 0.1 M, the L-isomer was used because of its greater
solubility (0.171 M) than the D, L-racemate (O.O7I M).
Imidazole was also recrystallized prior to use.
Tri-deuterio -D, L-leucine, which had been prepared by
Larry Levine (1963) by racemization of leucine in D^O
catalyzed by pyridoxal and alum, was analyzed by mass specteoscopy (Junk and Svec, 1964) as 97^ alpha-deuterated, including
92^ alpha-D-beta, beta'-Dg-leucine (or D^-Leu).
Sodium alpha-ketoisocaproate (KIC) was prepared from the
commercially-available keto acid, recrystallized, and char-
76
acterlzed by melting points of KIC and its 2, 4-dinltrophenylhydrazone, according to the procedures of Melster (1953).
It was used for standard curves in the analysis of KIC by the
quinoxallne, dlnltrophenylhydrazone, and quinolylhydrazone
methods.
Preparation of solutions
Reaction solutions and stock solutions of amino acids and
pyridoxal analogs were prepared with boiled, redistilled
water.
Amino acid stock solutions, 0.05 to O.167 M for
leucine, for example, usually had to be heated or dissolved
in acid or base.
Amino acid was carefully weighed into a
volumetric flask and dissolved by heating on a steam cone.
After cooling to room temperature the solution was diluted to
final volume. Solutions were filtered to remove any insoluble
matter and usually stored in a freezer at -20°C to keep for
an extended period of time.
Stock solutions of pyridoxal
analogs were quite stable when stored frozen, as determined
by spectral analyses.
It was found convenient to store these
solutions in small vials, to enable thawing of a portion and
warming to room temperature before use.
Concentrations of pyridoxal analogs were checked spectrophotometrically by comparing observed molar absorbancy indices
to values reported in the literature. (Refer to Table 1,
Review of Pertinent Literature, for spectral data of cationic,
dipolar ionic and anionic forms of pyridoxal analogs.)
Solutions for spectral analysis were 10"^ M PL, PLM, DPL, PLP,
77
APL, CPL in 0.1 N HCl; PL, PLM, DPL, PLP in 0.1 M phosphate
buffer -pH7, and in 0.1 N NaOH.
Spectrophotometrio Measurements of the Ionic
Equilibrium of Pyridoxal-N-methochloride
Absorbancies of prepared solutions of varying pH
Solutions were prepared by 1 to 5 dilutions of an aqueous
stock solution of 4.84 x 10~^ M pyridoxal-N-methochloride
(PLM) with acetate buffers of varying pH near the pK^^ in­
cluded. These solutions were 0.10 M in ionic strength,
assumed mainly due to buffer ions. Solutions of the same
concentration of PLM were prepared to be 0.01 N in HCl, pH
2.3, and 0.1 N in carbonate buffer, pH 10.1. Absorbancies
of each of these solutions were determined and blank
corrections were made at the three wavelengths of maximum
absorbancy from the recorded spectra (See Table 4, Figures
11 and 12 - Results and Discussion). All measurements were
made at 25°C.
Calculation of pK'g.
Calculation of the apparent pK'^ value was based on the
following assumptions: The spectrum of the cation form of
PLM is that obtained from the pH 2.3 solution (in agreement
with that of a pH 1 solution; compare a^ values in Table 4,
Results and Discussion); the spectrum of the dipolar ion form
is that of the pH 10.1 solution.
The fractions, f^, of PLM
78
In the conjugate base or dissociated form in solutions x, of
pH near the pK' were related to absorbancies of solutions x
and of solutions of 100^ cation form, f = 0, and of 100^
dipolar ion form, f = 1, by equation (l).
(1) f^ = (
)/( A^ - A^ )
The apparent pK'^ was then calculated from equation (2),
relating the equilibrium constant for dissociation of the
cation form with experimental pH values of solutions and
calculated fractions of dipolar ion form in these same
solutions.
(2) pK'a = pH - log
( f/l - f )
If the absorbancy of the 293.5 mp band is proportional
to the concentration of cation form, the pH at which the
absorbancy becomes an average value of the limiting values
(at pH 2.3 and at pH 10.l) is the pK' . The same assumption
a.
was made for the absorbancies of the 323 and 254.5 m)j bands
for the dipolar ion form. This method for calculation of
ionization constants from spectrophotometrlc data has been
described (Irvln and Irvin, 1947; Lunn and Morton, 1952;
Williams and Nellands, 1954; Metzler and Snell, 1955; Albert
and Phillips, 1956; Mason, 1958; Nakamoto and Martell, 1959b).
In general, for spectrophotometrlc measurements, the
direct proportionality between absorbancy (A) of a solution
with respect to the solvent at a certain wavelength and the
molar concentration (c) of the absorbing species was assumed
to hold:
A = a be, where a is the molar absorbancy index and
m
m
79
and b is the optical path length of the solution In cm.
Spectrophotometrlc Determination of the Formation
Constant for the Hydrogen-bonded Aldimine of
Pyridoxal-N-methochloride and Valine
Preparation of stock solutions
A stock phosphate buffer solution was prepared by
dissolving a calculated amount of dry potassium dihydrogen
phosphate in a calculated volume of standard 0.1 N NaOH in
a volumetric flask and diluting to the mark.
Ten stock
solutions (50 ml. each) were prepared by dissolving carefully
weighed amounts of dry D, L-valine in calculated volumes of
standard 0.1 N NaOH and adding 25 ml. of stock phosphate
buffer. (The dissolving of valine was hastened by heating
on a steam cone and shaking.) These solutions contained
amounts of phosphate, base and valine that were calculated
so that the final solutions for study of imine equilibria
would be 0.5 M in ionic strength, 0.1 M in phosphate, pH
6.0 and in ten graduated concentrations of valine convenient
for plotting reciprocal valine concentrations.
The concentration of stock pyridoxal methochloride
solution, prepared as 4.165 mg./ml., was determined spectrophotometrically at 293 mp to be 4.053 mg./ml. First, the con­
centration was calculated as O.OI86 M from the absorbancy of a
1 to 500 dilution of this stock solution in 0.1 N HCl, using
7.22 X 10^ for the molar absorbancy index of the hemiacetal
80
cation at this wavelength (Table 4, Results and Discussion).
The latter concentration of the pyridoxal methochlorlde
determined as the hemlacetal cation was used in calculations
of molar absorbancy indices involved in the measurement of
Imine equilibria.
Procedure for measuring imine equilibria
Solutions for measuring varying extents of imine for­
mation were prepared by adding 0.5 ml. of pyridoxal metho­
chlorlde (0.0186 M) to each of ten 25 ml.-volumetric flasks
and diluting to the marks with each of the ten respective
phosphate-buffered solutions of graduated valine concen­
trations. The average pH measured for the ten solutions was
6.02;^0.01.
These solutions were allowed to reach a state
of equilibrium in the formation of the hydrogen-bonded
aldimine of valine and pyridoxal methochlorlde.
Thirty
minutes was found to be a sufficient period of time for this
equilibration at 25°C, and the solutions were protected from
light to prevent loss of pyridoxal methochlorlde.
The
absorbancy of each solution was read Immediately after its
thirty-minute equilibration period against an appropriate
blank solution at 419
a wavelength of maximum absorbancy
of the hydrogen-bonded pyridoxal methochloride-vallne imine
(Table 2). Each blank solution had the same composition as
its corresponding imine solution, except that an equal volme
of water was substituted for the pyridoxal methochlorlde
81
solution.
Determination of molar absorbancy Index of pyridoxal
methochloride dipolar ion (Figure 14, Pt) at 419 ni)i
The concentration of a pyridoxal methochloride solution
(10 mg./ml.) was determined spectrophotometrlcally to be
0.00473 M (from the absorbancy at 293 mp of a 1 to 10 dilution
of this solution in 0.1 N HCl). From spectrophotometrlc
measurements at 4l9 mja of a 2 to 5 dilution (in 0.5 M phos­
phate, pH 6.00, 0.5 M ionic strength) the molar absorbancy
index, ap, of the PLM dipolar ion under these conditions was
calculated to be 98.8.
Derivation of an equation relating the equilibrium
constant with observed molar absorbancy indices
Consider the following net equilibrium for the formation
of the hydrogen-bonded aldlmlne of pyridoxal methochloride
and valine:
K.'
P± + v± ^
>.
PVi, where P-, V:, PVi are dipolar ions of PLM,
val, and their imlne, respectively. Assume that all of PLM,
val and their imlne are of dipolar ionic forms at pH 6.0.
This assumption may be justified since pK'
cL of PLM is 4.05,
pK'n of val is about 2.3, pK' of val is about 9 . 6 and pK'
^
3.
of PLM-val is 8.0 (Johnston et al., 1963). Let (P), (V), (PV)
represent equilibrium concentrations of PLM, val and their
imlne, respectively. The experimental conditions are such
that (V)»(P) or (PV).
The equilibrium constant for formation of the hydrogen-
82
bonded imlne may then be expressed by equation (3).
(3)
= (PV)/(P)(V) = 1/(V) • (PV)/(P)
Let the fraction of free (unreacted) PLM be expressed as f^
- (P)/[(P^ •*'CPV)], and the fraction of PLM reacted be expressed
as fp^ = (PV)/[(P) ^(PV)].
Then
' = l/(v) • f yf
or from
^pv ~
(4.) K^' = 1/(V) • ( 1 - f )/f
P
P
These fractions may also be related to molar absorbancy
Indices, as in equation
(5) a^=af+a f , where a Is the observed molar
0
p p
pv pv
o
absorbancy index at 4l9 mp. of a solution of P, V, and PV at
equilibrium^ a^^ and a^ are molar absorbancy indices of PV
and P, respectively, at 4l9 mvi.
Equation (5) may be simpli­
fied, in the same manner as equation (4), to obtain equation
% = Vp' V ^
After subtracting a^ from both sides of equation (6), solving
equation (4) for ( 1 - f^ ), and substituting its equivalent,
Kj_' (V) fp into equation (6) one obtains equation
(7)
- ap = (
(V)
Since fp = 14 + (PV)/(P)], and K ' (v) = (PV)/(P), from
equation (3), then fp = l/[l + K^'(V)].
Substituting this
expression for fp into equation (?), taking the reciprocal of
both sides of the resulting expression, and rearranging terms,
one obtains equation
(8) l/(a^ - a^) = 1/(V) • 1/K^.
- a^) *
- a^).
83
A graphical plot of l/( a^ - a^ ) vs. l/(v), as ordinate
and abscissa, respectively, yields.a straight line with slope
of 1/K.' (a
- a ) and Intercept of l/( a
- a ). Thus,
1
pv
p
pv
p'
K^' may be calculated as the ratio, intercept/slope. Equation
(8) is entirely analogous to the form of the equation of
Ketelaar et al. (1951, 1952), as expressed in Review of
Pertinent Literature..
General Procedure for Nonenzymlc Transamination in Model Systems
Appropriate amounts of stock solutions of amino acid and
pyrldoxal analog were pipetted into volumetric flasks, and
buffer, metal salt solutions or other additional catalysts
were added, if desired, before dilution.
The time of dilution
and mixing of reaction solutions was considered as time zero.
Aliquots of the reaction solutions were analyzed for keto
acid and pyrldoxal analog or measured spectrally at appro­
priate times.
Control solutions in which the amino acid was
omitted were used to check stability of pyrldoxal analogs
under reaction conditions, so that corrections could be
applied to any changes observed in the concentration of
pyrldoxal analog in the absence of amino acid if necessary.
The solutions were transferred to screw-capped test tubes
and kept in a constant-temperature water bath at 25°C during
the reactions; however, the method of Metzler and Snell
(1952b) was used for the preliminary comparisons carried out
near 100°C (Johnston et
, 1963).
Protection of all solu­
84
tions of pyridoxal analogs from photodeoompo.sltion by lab­
oratory (fluorescent) light was afforded by aluminum foilwrapping or by the water bath made opaque to visible light of
shorter wavelengths with a soluble dye (amaranth or aniline
red).
Water of the same temperature was circulated around
the cell compartments of spectrophotometers, and all pH and
absorbancy measurements were made at 25°C (in a room kept at
relatively constant temperature and humidity).
Analytical Methods
Dinitrophenylhydrazone method for keto acids
A diluted aliquot containing from zero up to two micromoles of keto acid and less than one micromole of pyridine
aldehyde were reacted with a reagent containing five micromoles of 2,4-dinitrophenylhydrazine and two millequivalents
of acid.
After thirty minutes of formation of the hydrazones
of keto acid and pyridine aldehyde, the pH was adjusted to
neutrality with two milliequivalents of base contained in a
phosphate buffer.
At neutral pH the hydrazones of pyridine
aldehydes have no net charge and were filtered off as red
precipitates. The filtrate contained the anionic keto acid
hydrazone. Extraction of this filtrate with toluene removed
excess reagent and any unprecipitated neutral hydrazones or
unreacted pyridine aldehyde. Equal volumes of the extracted
filtrate and 2.5 N sodium hydroxide were mixed and an uncharacterized, red colored product, having an absorption maximum
85
at 520 mp, was allowed to form for ten minutes before measuring
the absorbancy against a reagent blank treated in the same
manner.
More specific details of reagents and procedure of
this method developed by Metzler and Snell can be found in
their 1952 article {1952b).
This method for the analysis of keto acid formed in model
transamination systems of pyridoxal and amino acids was used
for following reactions at 100°C, where keto acid was rapidly
formed (within a few minutes of reaction) and pyridoxal was
rapidly converted to pyridoxamine, and for preliminary
studies of the same reactions at 25°C.
However, this method
was found inadequate for the sensitivity desired. In reactions
followed at 25°C, a much slower rate of keto acid formation
required a method specific for measuring small amounts of keto
acid in the presence of a large excess (about twenty times)
of pyridine aldehyde. In this situation analyses for keto
acid were up to 40 too low because of co-precipitation of the
dinitrophenylhydrazones of keto acid and pyridine aldehyde
(Metzler and Snell, 1952b). Another disadvantage of the
method was the cloudy suspension which sometimes formed in
the final color-developed samples. These two disadvantages
of this method resulted in unreliable reproducibility and
undermined the precision required for analysis of small
quantities of keto acid formed in these model systems.
The dinitrophenylhydrazone method using perchloric acid,
in place of sodium hydroxide, was considered too complex to be
86
useful, although it was used for analysis of the nonenzymic
transamination of alanine with pyridoxal at 25°C, which was
followed up to about 6fo completion of the reaction (Banks et
al.J 1961).
Instead, further modification of the Metzler-
Snell procedure was attempted.
Aliquot size was reduced,
dilution was eliminated and volumes were adjusted for con­
venience in calculation and for saving time.
Since it was
found that formation of dinitrophenylhydrazone of ketoisocaproate was complete in five minutes but slow and incomplete
in the case of pyridoxal, it was reasoned that dinitrophenyl­
hydrazone formation of pyridoxal could be held to a minimum
if only five minutes were allowed for this reaction.
The
problem of co-precipitation of a small amount of keto aciddinitrophenylhydrazone with a relatively large amount of
pyridoxal-dinitrophenylhydrazone was solved, because unreacted
pyridoxal could still be extracted from the neutralized
solution with toluene.
The cloudy suspension of toluene was
avoided by allowing for complete separation of layers after
the extraction, by carefully preventing toluene from getting
into the sample pipetted from the lower layer, and by shaking
the alkaline solution of keto acid-dinitrophenylhydrazone
vigorously, to enable dissolved air to force suspended toluene
out to the surface of the liquid.
The detailed procedure of this modified method is as
follows: 1)
Pipet sample containing less than four micro-
moles of keto acid and less than one micromole of pyridoxal
87
analog into one ml. of dinitrophenylhydrazine reagent in a 10
ml.-volumetric flask. Reagents are prepared as usual (Metzler
and Snellj 1952b). 2) Allow just five minutes for reaction.
3) Add five ml. basic-buffer reagent, dilute to the mark and
mix well. 4) Filter into a 30 ml.-or 1 oz.-bottle containing
about 10 ml. toluene; cover with ground-glass stopper and
shake well. 5) Allow layers to separate, before carefully
taking five ml. of aqueous layer, avoiding toluene in pip­
etting and transferring the aliquot to a test tube. 6) Add
five ml. 2.5 N NaOHj cover tube with screw-cap and shake
contents vigorously. 7) Measure A^^^
with spectrophoto­
meter, after ten minutes of color development, in one-cm,
calibrated, pyrex cells, against a reagent blank treated in
the same manner.
Although this modified procedure was believed to have
made several improvements, this method for keto acids was
still lacking in the desired reproducibility and precision.
Results depended too much upon the conditions of the reaction
solution being analyzed, such as the nature and concentration of
buffer, metal ion, pyridoxal analog and pH. It was concluded
that the method was too complex, due to the lack of specificity
of the reagent toward keto acids and that the colored product
was too time-sensitive (faded gradually).
Others have discussed
the following difficulties and disadvantages of the method:
possible loss of keto acid-dinitrophenylhydrazone by ex­
traction, acid-dependency of the reversible reaction between
88
keto acid and dinitrophenylhydrazine, and instability of the
colored product (Meister and Abendscheln, 1956; Robins et al.,
1956; Spikner and Towne, 1962).
Quinoxallne method for alpha-keto acids
A specific method for the fluorometric microdetermination
of alpha-keto acids (Spikner and Towne, 1962), was tried. The
procedure for reacting ortho-phenylenediamine in 50^ sulfuric
acid with an aliquot of alpha-keto acid in the presence of
other carbonyl compounds was simple and convenient. However,
a heating period was required for the condensation.
quinoxaline product was very stable.
The
But the unavailability
of a reliable photoflourometer prevented the use of this
method.
A spectrophotometric method for the specific o-phenylenediamine reagent was used for analysis of keto acid produced by
a model transamination system (Bruice and Topping, 1963a).
But the procedure called for isolation of the crystalline
derivative, after a 24-hour reaction, before making a solu­
tion for spectrophotometry. The procedure lasted two days
before results could be obtained.
Ethanolimine method for pyridoxal analogs
Analysis of the remaining amount of pyridoxal analog in a
model transamination system was made by a method developed pre­
viously (Metzler and Snell, 1952b). An aliquot containing up
to four micromoles of pyridine aldehyde was added to five ml.
89
50^ ( b y volume) ethanolamine (EOA), and the mixture was
diluted to ten ml. The condensation product forms almost
immediately, is stable for at least an hour, and can be
spectrophotometrically measured at absorption maxima char­
acteristic of the corresponding imine anions (pH 11). These
maxima are 365 mu(a = 6.82 x 10^ )for pyridoxal, 375
'
m
mjui ( am r 5.38 x 10^ ) for pyridoxal methochloride, 350 mp.
( a^ = 3.85 X 10^ ) for pyridoxal phosphate, and 344 mpL
( a^ = 4.67 X lo3 ) for deoxypyridoxal. A molar absorbancy
index of 6.74 x 10 has been reported for the ethanolimine of
pyridoxal at 362.5 np and was constant over a temperature
range of I8 to 27°C (Banks et al., 1961).
Although this was a simple, convenient and accurate
method for measuring amounts of pyridoxal analog plus its
amino acid aldimine remaining in a reaction solution, no
reaction could be detected with "5-carboxypyridoxal" or with
"alpha^-pyridoxalylacetic acid," the carboxylate analogs of
pyridoxal, and ethanolamine (0.2 ml. 1 mM analog/5 ml. 25^
EOA gave no spectral change or peak above 330 mp). In fact,
the spectrum of "carboxypyridoxal" did not change much from
an aqueous solution of pH 6 to an ethanolamine solution of
about pH 11 ( max, from 319 to 315 mp and a^ from 7.0 to
9.5 X lo3 ).
Alkaline-acetone method for pyridoxal (Siegel and Blake, 1962)
A red condensation product from acetone and pyridoxal in
90
alkaline solution formed slowly, but this method specified the
measurement of absorbancy at 420 mp at the point of greatest
change in absorbancy with respect to time: fifteen minutes
after mixing. This absorbancy increased at a much slower rate
after thirty minutes at room temperature.
The claims for
simplicity, convenience and accuracy of this method were
greatly exaggerated (Siegel and Blake, 1962).
It has been
found that the instability of the colored product, the
decreased sensitivity to pyridoxal, and other undesirable
features make this method inferior to the ethanolimine method
for the analysis of pyridoxal.
Quinolylhydrazone (QH) method for both pyridoxal analogs and
alpha-keto acids
A modified method for the microdetermination of alphaketo acids by a quinolylhydrazine (QH) reagent, originally
described by Robins et al., (1956), was found to be very
simple, accurate, reproducible and suitable for not only
detecting keto acid formation, but also for detecting the
decrease in concentrations of pyridoxal analogs in a model
transamination system.
Alpha-keto acids react with 3-hydra-
zinoquinoline (QH) to form a specific hydrazone of unknown
structure (Unknown, because beta-and gamma-keto acids,
aldehydes and ketones form different or unstable products
with absorbancies at 305 mp which virtually disappear in time
or upon dilution.), having a very high absorption maximum at
305 myi in acid solution(a
- 24.3 x 10^ for KIC ). This QH
91
reagent has been shown to be effective for measurement of as
little as 0.4 millimicromoles of 13 different alpha-keto acids
(Robins et al., 1956).
Concentrations of both keto acid and pyridoxal analog may
be determined in the same QH solution.
Absorbancy changes at
absorption maxima of 305 and 400 m)i are linearly related to
concentrations of QH products of keto acid and pyridoxal ana­
log, respectively, e. g., alpha-ketolsocaproate and pyridoxal
phosphate (Figure 8). Pyridoxal phosphate and pyrldoxamine
have identical molar absorbancy indices at 305
(4.48 x 10^)
In 0.01 N HCl. This would have to be true for other pyridoxal
and pyrldoxamine analogs in order for the QH method to be
quantitatively useful In these analyses of reaction components.
These standard curves are compared with the ethanolimine
method for pyridoxal phosphate, which is about 1/5 as sensitive
as the QH method for PLP.
Agreement of the EOA and QH methods
in following the decrease in concentrations of pyridoxal phos­
phate In a nonenzymlc transamination with leucine is substan­
tiated in Results and Discussion. Roze (1964) has also adapted
the QH method for the analysis of pyridoxal phosphate (a m ~
23.2 X 10^ at 400 m;i).
The original QH method of Robins jet
, (1956) was
modified for the analysis of model transamination reactions
by decreasing the size of aliquots and reagent proportionately
and by following decreases in absorbancy at 400 myi correspond­
ing to pyridoxal analog.
An aliquot of model system solution
Figure 8. Standard curves for analysis of alpha-ketoiisocaproate (KIC) and
pyridoxal phosphate (PLP) by the quinolylhydrazone (QH) method and
of PLP by the ethanolimine (EOA) method
Absorbancies vs. milliraolar concentrations of these components in
solutions of 1.0 mM leu, 0.20 mM PLP + KIC or PM, such that (KIC) =
(PM). Over-all dilutions of 1 to 10 were made in EGA or QH
reagents, and further 1 to 10 dilutions of these QH solutions were
made in 0.01 N HCI after complete reaction.
Sensitivity, mM/A
To calculate mM
0 A vs. (PLP) by the EOA method
0.257
2.57 A/x
• A vs. (PLP) by the QH method
0.0572
5.72 A/x - 0 . 6 3 / x
A A vs. (KIC) by the QH method
0.0471
4.71 A/x - 0.57/x
(x
is aliquot size in ml.)
—1
0'5
1
91
1
9'\
1
Vl
r
9b
—I
T"
9b
h'o
3b/
%
'0
30
a
cb S
o
u
03
OOhj^
ro
cr\
Vo z
X
QOCy
Qb
009
ob
^b
J
L
J
1
L
i
1
94
containing a total of about two micromoles of pyridoxal analog,
amino acid Imine, and keto acid was added to 5 ml. of 1.0 mM
OH in 0.02 N HCl, and was diluted to 10 ml. This QH reagent
was reported to be stable for two to three weeks at 4°C, and
was prepared by diluting a stock solution 1 to 10. The more
concentrated QH stock solution (232 rag/100 ml. or 10.0 mM QH
in 0.20 M HCl) was kept frozen at -20°C, where it was reported
to be stable for about l8 months (Robins^
, 1956). The
final QH .concentration of 0.5 mM will be about 2^ times in
excess of the amount needed to completely react with model
system components.
At room temperature the time for complete reaction may
be determined directly by following the increasing absorbancy
at 280 mji of the reaction solution, as suggested by Robins
et a^. (1956), or by following the increasing absorbancy at
305 m)j of aliquots of the reaction solution diluted 1 to 10
in 0.01 N HCl. The time for complete QH reaction was longer
in the presence of buffer. It took 90 to 110 minutes for the
absorbancy at 305 m}i to reach a maximum, if 1 M buffer was
present in the original model system solution (or about 0.02
M buffer in the QH reaction in HCl). But it took only 50 to
70 minutes for the absorbancy at 305 myi to reach a maximum,
if the original model system solution was unbuffered.
After
completion of the QH reaction, the maximum absorbancies at
305 and 400 m;i are measured for samples diluted 1 to 10 in
0.01 N HCl, which are stable for at least two hours. Fresh
95
OH reagent was treated In the same manner as the samples and
used as an absorbancy blank.
An ultraviolet lamp was found to
be convenient and appropriate for absorbancy measurements at
both 305 and 400 mp.
Direct spectral analysis
Model system reactions have also been followed by di­
rectly measuring the absorbancies of the reaction solutions
at appropriate wavelengths. For solutions with high reactant
concentrations calibrated quartz cell spacers (8.0, 9.0, 9.5,
9.9 mm) have been used to decrease the optical path length
(to 2.0, 1.0, 0.5, 0.1 mm, respectively) as required. Loss
of aldimine has been followed at wavelengths of maximum
absorption between 400 and 425 mjn, depending upon both the
pH and the pyridoxal analog used. Increases in absorbancy
between 320 and 335
have been considered to be due to
formation of ketimine and/or pyridoxamine analog after the
rate-determining step.
Other electronic absorption maxima
below 300 nju changed during a reaction (increased and/or
decreased), but it was not understood what molecular changes
were responsible for these spectral changes. Furthermore,
these changes were not as pronounced as were the other two
absorption bands.
Chromatography of reaction solutions
Alpha-ketoisocaproate (KIC) and pyridoxal analog in a
portion of reaction solution treated with 2,4-dinitrophenyl-
96
hydrazine (DNPH) were Identified as dlnltrophenylhydrazone
(DNP) derivatives by one-dlmenslonal, ascending chromatography
on paper and on thin-layer silica gel plates.
The thin-layer
chromatographic techniques were found to be most convenient
because of their rapid solvent development (3-4 hours) and
clarity of spots. It was found best to spot the reaction
solution directly on the chromatogram.
When this was fol­
lowed with the application of DNPH reagent to the spot,
yellow DNP-derivatives were produced, both with keto acid
formed in the nonenzymlc transamination and with the pyridoxal
analog remaining in the reaction solution.
After development these same chromatograms were sprayed
with a nlnhydrln solution to detect the constituents with
amino groups:
purple or pink spots were identified as amino
acids or esters, orange spots as pyrldoxamlne analogs. When
the reaction solutions were not treated with DNPH reagent,
but were sprayed with Glbbs' reagent and nlnhydrln after
development, blue or gray spots for pyridoxal and pyrldoxamlne
analogs, which rapidly faded to brown spots, were detected
along with the nlnhydrln-sensitive components.
Also, both
nlnhydrln and Glbbs' reagent reacted at the same large spot,
which could have been the aldlmlne of leucine and deoxypyridoxal, since a large brown spot with pink or purple edges
was found. Glbbs' reagent or N, 2, 6-trichloro-p-benzoquinonelmine in ammonia solution is specific for para-unsubstituted phenols (See Block et al., 1955^ p. 229, for prep­
97
aration of reagent and chromatographic procedures).
When this
reagent was applied to spots of the reaction solution before
development of the chromatogram, different R values were
P
obtained for brown, blue, and gray spots, but these spots were
more distinct and the brown background from the Gibbs'
reagent spray was avoided.
Solvent systems used were n-butanol (4), acetic acid (l),
water (5) and the upper layer of n-butanol (5)j ethanol (l),
water (4), where parts by volume are in parentheses.
Some of
these chromatographic techniques have been used previously
for qualitative identification of components in nonenzymlc
transamination reactions (Metzler and Snell, 1952aj Metzler
et
1954b).
Kinetic Methods
Typical graphs of rate data for the nonenzymlc trans­
amination of amino acids with pyridoxal analogs are presented
in Results and Discussion.
The absorbancles (at wavelengths
depending on the particular analytical method) of aliquots
of reaction solution were recorded at various times throughout
the time of the reaction, and have been previously shown to
be proportional to concentrations of a reactant or a product
(Figure 8). The rate was determined from these reaction
plots by one or more of the following methods.
l) Calculation of the initial rate, dP/dt, was made
from the slope of the tangent to the curve at time zero.
98
since the relationship between concentration of the absorbing
species changing with time and the measured absorbancy were
known. The value dP/dt was converted to a rate constant by
means of either the simple 1st or 2nd order rate expressions;
(9) -dP/dt =
(P) or = kg (P)(A), where (P) and (A) are
concentrations of pyridoxal analog and amino acid, respectively.
This method is not accurate in the determination of rate
constants for model transamination reactions because of the
uncertainties in the initial or instantaneous concentrations
of "effective" reactants, owing to equilibrium conditions
(i. e.j only a portion of the initial reactants are converted
to products at equilibrium).
2) Calculation of the relative rates, in comparing two
reactions from the ratio of inverse times required to complete
the same percent of reaction in each case, was made from the
intergrated rate expression:
log ( Px/Po) = k t = k t .
1 1
2 2
(10) k^/kg = tg/t^
This method is able to quickly compare relative rate constants,
and if one of the constants is already known from a previous
determination, it can be very useful.
3)
Calculation of the rate constant was made from a
fractional-life period, such as the half-time of reaction
from the integrated first-order equation:
t, - (in 2)/k ,
2 "
1
where k^ = first-order rate constant.
(11)
= 0.693/ti
The problem of determining t^ is complicated because concen2
99
trations of reactants at equilibrium need to be known, in
order to be able to accurately calculate the "effective" con­
centrations of reactants.
But the half-time has been found
graphically by following a reaction practically to equilibrium
and assuming an equilibrium value.
4) Calculation of rate constants from simple firstand second-order plots:
(12) slope of log (X) vs. t = first-order rate constant and
(13) slope of l/(X) vs. t = second-order rate constant, where
(x)
is some function of the concentration of a reactant
or product
The application of simple kinetic relationships has been
found to be inadequate in the complex sequence of reaction
steps which constitute the transamination reaction, proceeding
1 to
an equilibrium position.
But kinetic methods have been
^^ror - ' '" '
characterize such a reaction fairly
we:
of pyridoxal phosphate with leucine,
second-order rate constants were calculated by the usual
graphical method.
These reactions were followed up to 30^
completion.
However, reactions having pseudo first-order conditions
(excess amino acid) did not follow simple first-order kinetics.
(Reactions should be followed for at least 25 to 30^ of
completion, in order to be able to distinguish first-andsecond-order.) But when equilibrium values were determined
graphically by a method of successive approximations, the
100
"effective" concentrations of reactants could be calculated
from the known initial values and these equilibrium values.
A plot of log ( Xg -
- X^ ) vs. t, where X is concen­
tration or absorbancy (Figure 9) then gave a straight line,
from which the first-order rate constant could be calculated
from the relationship:
(14) k .
ODS
= -2.303 slope (in 10 = 2.303)
Calculation of pseudo first-order rate constants were
made from Guggenheim plots. The Guggenheim method takes equi­
librium conditions into account, without requiring measurements
of initial or final values (Guggenheim, 1926). It is Ideally
suited for the determination of pseudo first-order rate con­
stants for these nonenzymic transaminations.
The pseudo
first-order conditions, in a unimolecular reaction of pyridoxal analog with an excess of amino acid, the following
Guggenheim relationship may be derived:
•
Suppose A^, ...A, ...A^ are n readings of absorbancy
which are linearly proportional to the concentration of some
molecular species, which increases or decreases in time with
a rate directly related to the reaction rate of nonenzymic
transamination. If these absorbancy readings are taken at
times t^, ...t, ...t^, without any restriction as to inter­
vals, and n more readings A'^, ...A', ...A'^, are taken at
times t^ + T, ...t + T, ...t^
T, each a constant time
interval T after one of the previous set, then,|A^ - A|
=]Af - AQ| e~^^ andjA^ - A'|=|A^ - A^j
~
where A^
and AQ are final (equilibrium) and initial values, respec­
tively, and k is the sum of the rate constants of the two
opposed processes in this reversible reaction.
Figure 9-
First-order plots for the nonenzymic transamination between pyridoxal
phosphate and leucine: log (A - A )/( A - A ) vs. t, where A ,
Go
et
o
A^j A^ are absorbancy values at time zero, t and equilibrium,
respectively, using semi-logarithmic plot, where (A -A )/(A -A )
e,
o
e
t
ordinate is logarithm scale, but numbers are not logs
k
X 10^ sec"^
obs
O
•
vs. (PLP) by EOA method
20.4
vs. (PLP) by QH method
22.2
AA^Q^mp vs. (KIC) by QH method
12.7
A^ values found graphically by method of successive approximations to give best
straight lines in this plot ( T = 30 hours )
Pseudo first-order conditions; 10 mM PLP
50 mM leu
1.0 M acetate buffer, pH 4.6, 25°c.
T
T
T
T
I
I
I
1
1
I
-r
r
6
•10'-
9ôOy
r6-
5-
M
O
"Re-Ro
•Rg-'R-t
1.5
•10'
J
10
20
1
30
T I M E HRS.
1
1
40
I
I
50
I
LJ
60
103
Taking natural logarithms of both sides of the former
equation and solving for k, one obtains the complete Inte­
grated form of the first-order relationship for the reaction
which has an equilibrium value for absorbancy or concentration
greater than zero: k = logg |Af - Ag|/|A.^ - Aj. But combining
t
both of the above equations by substitution:
Ja' - A|-jA'^ - aJ (1 - e"^^)e"^^. Taking logarithms:
(15) kt + loggjA' - A I = loggjAf - AqICI - e'kt)
Since the right side of this equation is constant, plotting
logicjA' - AI vs. t gives a straight line with slope of -k
logioe; or
(16) k = -2.303 A log|A' - AI/At
There is no extrapolation and every observed reading is
used only once. The longer the constant time interval T, the
greater the value of |a' - Ajfor a given t and the more
accurate the method.
Provided that T is several times as
great as the time of half-completion of the reaction, the
accuracy will be of the same order as in taking n ordinary
I
readings and further taking n readings of the end-point
spread over an interval of time equal to that spent on the
ordinary readings; in fact, it is equivalent to obtaining
a very accurate end point and using It in the usual manner
(Guggenheim, 1926).
An analogous method for a second-order
reaction with equivalent concentrations has been developed
by Roseveare (Frost and Pearson, 1961, p. 150).
Comparisons of Guggenheim plots for two different nonenzymic transamination reactions, using different analytical
methods, are presented in Figure 10. Two examples are shown
for deviations from linearity or pseudo first-order con­
ditions, in measuring the loss of PLP. These deviations were
always concave downward, and two straight lines of different
Figure 10. Guggenheim plots for nonenzymic transaminations
of leucine with pyridoxal (PL) and with deoxypyridoxal (DPL), semi-logarithmic scale: log
Ax vs. t
mM:
change in millimolar concentrations of pyridoxal (PL),
as measured by the ethanolimine method, and of ketoisocaproate (KIC), as measured by the dinitrophenylhydrazone method
Points taken from plot of mM vs. t; T = 20 hrs.
Initial conditions:
10 mM PL
120 mM leu
2 mM alum
0.1 M acetate buffer, pH 4.2,
25OC
^obs ^ lO^sec"^
°^
dP/dt (2 hrs.) x 10®M/sec
dP/dt/5 x
10-3 M
PL
2.7
15
ca. 3
KIC
2.85
13
oa. 3
logAAbsorbancy vs. t;
T = 24 hrs.
R
-1
k
X ICrsec
obs
(initial (final
slope)
slope)
A
305 mja vs. (KIC) by QH method
2.80
2.80
•400 m;: vs. (DPL) by QH method
4.77
2.76
O 413 ïïip. vs. Decrease in (aldimine) by
direct spectral analysis
6.4
2.9
Initial conditions: 10 mM DPL
100 mM leu
1.0 M acetate buffer, pH 4.2, 25 C
105
AmM
PL
crô
20
TIME,H0UR3
106
slope were drawn through the points of the graph.
"Initial"
and "final" rate constants were determined from these slopes
(as usual, by equation (l6), and the initial values were as
much as 2 to 3 times higher than the final values.
These deviations were usually only observed for decreasing
concentrations of pyridoxal analog, and even then, were
observed for probably about half of the reactions studied.
Rates based on keto acid formation were generally considered
more reliable. For further discussion of kinetic and stoichio­
metric anomalies, see Results and Discussion. Rate constants
obtained by various kinetic methods are compared under the
same and different experimental conditions.
It was concluded
that the simplest and most reliable method was the Guggenheim
plot using log ( A' - A ) vs. t.
107
RESULTS AND DISCUSSION
Absorption Spectrum and Ionic Equilibrium of
Pyridoxal-N-methochlorid e
The absorption of pyridoxal-N-methochloride, in solutions
of varying pH in the range of pH 1 to 10, was measured in the
region of 220 to 400 m;i (Figure 11).
The molar absorbancy
indices at the three absorption maxima and at the two isosbestic points were calculated (Table 4).
These values
generally agreed with those reported by Johnston et
,
(1963).
The presence of isosbestic points in the spectra of pyri­
doxal methochloride solutions of varying pH indicated that
only a single equilibrium, involving two ionic forms, was
being affected and that it should be possible to calculate
the ionization constant directly from the spectrophotometric
data. The acid dissociation constant and apparent pK'
a
value were calculated from measurements of the absorbancies
at 293.5 and 323 mp at the three pH values nearest 4 (Figure
12).
The average of these six calculated pK'
values was 4.05,
and the average deviation from this mean value was ±0.03. The
corresponding values for the equilibrium or dissociation
-5
constant were: K' = 9.1 ±0.6 X 10 .
The other data were discarded because the calculated f
values were too high (>0.9). None of the data at 254.5 mp
108
Table 4. Comparison of absorption spectra of pyrldoxal-Nmethochloride and pyrldoxal, 25°C
PLM^
Ionic form®
i\max, mjj.
a^ X 10-3
m
S X 10"^ ixmax, mp.
Hemiacetal
cation (Aa)
pH 1^
pH 2.3
293
293.5
7.2
288
323
255
1:1
317
323.0
254.5
7.78
4.35
267
1.65
4.13
7.22
9.0
Dipolar ion (Ab)
pH 10^
pH 10.1
Isosbestlc
points
304
252
1:0
263
2.05
4.4
300
^Structures - Figure 1
^Source: Metzler and Snell, 1955
^Structures - Figure 2
^Source; Johnston et al., 1963; solutions of 0.5 M In
Ionic strength with NaCl, compared with experimental
values from solutions 0.1 M in ionic strength
(the secondary absorption maximum of the dipolar ion, 1. e.,
ft-If 2* transition) were used for determination of the pK^
because of interference from carbonate absorption.
Thus,
carbonate is not a good buffer to use for obtaining the absorb-
Figure 11. Comparison of absorption spectra of pyridoxalN-methochloride and pyridoxal
pyridoxal methochloride
_ _ _ _ _
pyridoxal
cation forms, pH 1 (HCI solution)
dipolar ion forms, pH 10 (carbonate buffer)
0.5 M ionic strength, 25°C
(Johnston et al., 19^3)
MOLFIR FlBSORBFlNCy INDEX ^ "R^ x IGT^
GQ^
Figure 12. Spectrophotometric titration curves, pH vs.
absorbancy of pyridoxal methochlorlde solu­
tions, constructed from calculated pK' and
equations (l) and (2), Experimental &
Experimental points for solutions of pyri­
doxal methochlorlde with pH controlled by:
•0.01 N HCI
O acetate buffers
A carbonate buffer
0.10 M ionic strength, 25°C, assumed mainly
due to buffer ions
112
07
Q6
Q5
[fl 03
02
293/5 nyj.
4
10
113
ancy of the dipolar ion ( f = 1 ) at 254.5 mji.
Comparisons with pyridoxal
Prom this spectral study of the equilibrium between the
two ionic forms of pyridoxal methochlorlde ( Aa, Ab - Figure
2, Review of Pertinent Literature), It was concluded that N-
methylation of pyridoxal results in slight, but quantitatively
distinguishable, changes in its properties.
The absorption
spectra of pyridoxal and its N-methyl analog were similar,
except that N-methylatlon of pyridoxal resulted in shifts of
the absorption maxima to longer wavelengths, by as much as
5-6 mji for the higher wavelength band (Johnston ^
, 1963;
Figure 11). This shift is in agreement with earlier obser­
vations of 3-hydroxypyridlne compounds (Metzler and Snell,
1955). Furthermore, the spectra of both pyridoxal and its
N-methyl analog exhibited bathochromic shifts (shifts of
absorption maxima to longer wavelengths) with increasing pH
(Figure 11).
This "red-shift" caused by the ionization of
the phenolic hydrogen of pyridoxine analogs has been observed
by others (Heyl jet al., 1951; Metzler and Snell, 1955;
Nakamoto and Kartell, 1959a; Williams and Nellands, 1954).
The apparent pK' value of 4.05 for pyridoxal methoa
chloride corresponds with the value of 4.20 for dissociation
of the phenolic group of pyridoxal (Table 1, Review of Pertinent
Literature).
This lower pK' value for the N-methyl analog
a
may be a reflection on the different relative proportions of
114
hemlacetal to free aldehyde.
A higher fraction of free
aldehyde (1.7 times higher) has been estimated for pyridoxal
methochloride (Johnston et al., 1963), as compared to pyri­
doxal under the same conditions.
The aldehyde group, an
electron-attracting substituent, is expected to enhance the
acidity of the phenolic group. Electron-attracting inductive
effects of a few other substituent groups on the ionization
I constants of 3-hydroxypyridine-4-aldehydes, in particular
(Nakamoto and Martell, 19591») and of pyridoxine analogs, in
general (Williams and Neilands, 1954), have been discussed.
Significance of ionization constants of pyridoxal analogs
Dissociation constants of pyridoxal analogs have been
considered especially useful in any interpretation of pH
effects on their reactions with amino acids. These spectrophotometric methods used for the study of ionic equilibria
of pyridoxal-N-methochloride should be applicable to other
pyridoxal analogs, as they are synthesized for studies on
For example, pK' values of 3.96 and 8.4 were
a
spectrophotometrically determined for "alpha^-pyridoxalyla-
model systems.
cetic acid" by P. Scott Furbish (See Table 1 for complete
private communication).
Attempts to study mechanisms of reactions of pyridoxal
analogs with amino acids spectroscopically in model systems
have led to involved but necessary studies of the absorption
spectra of hydroxypyridinealdehydes as functions of pH,
115
because of the complicated equilibria of pyridoxal analogs in
solution (Nakamoto and Kartell, 1959&; Martell, 1963).
Furthermore, detailed understanding of the biological function
and mechanism of action of vitamin Bg coenzymes requires know­
ledge of their ionization constants of acidic groups and the
relative amounts of various ionic forms present in solution.
This information can be used to interpret pH effects on the
binding of these coenzymes to apoenzymes and to substrates
(Williams and Neilands, 195^; Metzler and Snell, 1955;
Snell and Metzler, 1956; Martell, 1963).
Hemiacetal formation
The ratio of hemiacetal dipolar ion to free aldehyde
dipolar ion was estimated to be 46 for pyridoxal methochloride, compared to a value of 80 for pyridoxal (Johnston
et
, 1963).
An explanation for this difference is based
upon the relative electromeric or electron-donating, inductive
effects of a methyl group and a hydrogen atom bound to the
pyridine ring nitrogen atom of pyridoxal.
This effect is
generally larger for a methyl group than a hydrogen atom and
therefore the positive character of the charge on the nitrogen
atom is less for pyridoxal methochloride than for pyridoxal,
other factors being equal.
This difference in positive character of nitrogen atoms
is reflected in a difference in mesomeric or resonance effects
of the pyridine ring on the positive character of the formyl
116
carbon atoms of these two compounds.
It follows that this
mesomeric effect is less in the case of pyridoxal, leading to
a greater positive character on its formyl carbon atom, and
thus, a greater electrophilicity for its formyl carbon atom
toward the nucleophilic hydroxymethyl group in the adjacent
position on the ring.
This greater relative amount of free
aldehyde in pyridoxal methochloride than in pyridoxal should
be considered, along with other effects of N-methylation or
protonation, in interpreting their relative reaction rates
with amino acids.
Equilibrium Constant for the Formation of
Hydrogen-bonded Aldlmlne of
Pyridoxal-N-methochlorlde and Valine
Although the extent of hydrogen-bonded aldlmlne for­
mation from pyridoxal methochlorlde and valine was estimated
(See Review of Pertinent Literature, Johnston ^
, 1963),
confirmation by an independent and more quantitative method
was made.
Relative amounts of imlne were determined spectro-
photometrlcally from a series of solutions of varying valine
concentration from O.O8 to 0.5 M and of constant pyridoxal
methochlorlde concentration at pH 6.0 (Table 5). The graph­
ical method of Ketelaar et aj., (1951^ 1952) was used to
calculate the molar absorbancy index and the formation con­
stant for the hydrogen-bonded imlne dipolar ion (Pc-d, Figure
4). The best experimental data (Table 5) for the equilibrium
117
under consideration are graphically depicted in Figure 13.
The values for a , K.' and log K.' from different methods
pv' 1
1
are compared in Table 6. Application of the method of least
squares to the data was considered to yield the most reliable
values, and resulted in a calculated log K^' value of -0.02
±0.1. It was noted that the relatively small changes in the
values for the slope and extrapolated intercept in a plot of
10V( Bg - a^ ) vs. 1/(V) led to unexpectedly large changes
in the values calculated for a^^,
' and log
'. These
values were calculated from equation (8), Experimental,
where K^' = intercept/slope and a^_^ = 1/lntercept 4- a^, from
the plot of the data of Table 5 in Figure 13.
Within experimental error and the error of the graphical
extrapolation, the formation constant for the hydrogen-bonded
Imlne dipolar ion of pyrldoxal methochloride and valine
(log K^' = -O.IO) calculated by an independent method was
verified.
At pH 6, about 30^ of the PLM is converted to the
valine imlne, under these conditions. The graphical method
may be more convenient than the method of successive approx­
imations (Johnston £t ^., 1963), but it is not more accurate.
The dissociation of the hydrogen-bonded proton from this
dipolar ionic form is governed by pK^^y = 8.0, calculated by
a successive approximation method (Johnston et al., 1963).
All equilibrium constants are apparent constants expressed in
terms of molar concentrations, except,that apparent hydrogen
ion activities (from pH - meter readings) were used instead of
118
Table 5. Data for the determination of the formation constant
for the hydrogen-bonged imine of pyridoxal methochloride and valine '
mxi
^0 X
10^
vs. blank 10-3 ;X 10-3 (a -a ;
0
p
Graph
points
(V), M
1
0.5000
2.000
0.885
2.375
2.276
0.4396
2
0.4000
2.500 •
0.745
2.000
1.901
0.5260
3
0.3330
3.000
0.648
1.739 1.640
0.6098
4
0.2500
4.000
0.515
1.383
1.284
0.7903
5
0.2000
5.000
0.431
1.157
1.058
0.9450
6
0.1670
6.000
0.377
1.012 0.9132 1.095
7
0.1428
7.000
0.330
0.8860
8
0.1250
8.000
0.292
0.7839 0.6851 1.460
9
0.1000 10.000
0.245
0.6579 0.5591 1.789
10
0.0833 12.000
0.213
0.5719 0.4731
ro
0
A419
1.271
2.113
^Experimental conditions were 0.1 M phosphate, pH 6.00
^0.01, 0.500 M in ionic strength, 25®C
^Symbols used in the headings for this table are: (V),
valine concentration; A, absorbancy; a^, molar absorbancy
index for solutions of hydrogen-bonded pyridoxal methpchloride-valine imine, PVj^, and a = A^^q/3.726 X 10"^ M
(initial concentration of pyridoxal metnochloride; a ,
molar absorbancy index for pyridoxal methochloride
dipolar ion, P^, at 4l9 r#
®In Figure 13
Figure 13. Formation of hydrogen-bonded imine of pyridoxal
methochloride and valine as a function of
reciprocal valine concentration, 30 min. after
mixing in the dark, pH 6 phosphate buffer,
0.5 M ionic strength, 25°C (Plot of data of
Table 5:
^ .
10VC
- Sp ) vs. 1/(V)
Possible slopes and intercepts (See Table 6)
Maximum slope:
straight line through first 4
points
Minimum slope:
straight line through points
4,5,6
Average of minimum and maximum slopes
Least-squares slope:
straight line through all
ten points calculated by
method of least squares
120
1
T
r
I
I
I
I
r
HO
VHO
Y
CHj
CH3
PV±
(oo-ap)'
J
J
L
4
w
6
M"
I
L
8
10
I
I
L
12
121
Table 6. Comparison of results from spectrophotometrlc
measurements of the formation of hydrogen-bonded
aldimine of pyrldoxal methochloride and valine,
calculated by different methods
Intercept
X 10-
Method
Slops
X 10-
Straight line through
1st 4 points^
0.1753
0.087
Straight line through
points 4,5,0.1526
0.179
Ave. of graphical
extremes^
0.1639
+0.0113
Least squares method
applied to data"
0.l6l2
0.133
+0.046
0.1541
Indirect estimate^
Best line fitted to
points by eye"
0.168
0.105
a.
pv
X 10-
Kl'
log
K,'
7.62 0.812 -0.09
+0.36 +0.26
6.6
0.954 -0.0205
+8
+0.3 +0.1
7.5
0.788 -0.10
9.6
0.624 -0.205
^Figure 13
^Table 5
"^Source: Johnston et al., 1963
^Reported as note in Johnston ^ aj., 1963; a values of
imine solutions were determined at 419 nip, n8t at 319 mp
(typographical error in publication)
122
1
hydrogen ion concentrations.
Comparisons with imine formation constants for pyridoxal and
valine
Furthermore, it can be concluded that pyridoxal methochloride and pyridoxal are quite similar in the extent of
imine formation with valine, except for the lack of an equi­
librium for the imine dianion,
in the case of the former
(Figure l4). The relative log K^' values indicate that the
PLM-valine imine (log K^« = -0.02 ±0.1) is just slightly more
stable than the corresponding pyridoxal imine (log K^' = -0,27).
Also, from calculated values of log K^" (Figure l4), the
hydrogen-bonded anionic imine from pyridoxal is approximately
100 times more stable than the non-hydrogen bonded form from
pyridoxal methochloride.
Hydrogen-bonding lends stability
to the imine, because the equilibrium constant for transfer
of the proton from the ring nitrogen to the hydrogen-bonded
position was estimated to be 125 (Johnston et
, 1963).
Yet in pyridoxal itself the tendency is for the proton to
stay on the ring nitrogen, with K = 0.l4 for transfer of the
proton to the phenolate anion position. (For comparison with
enzyme spectra and ionic equilibria see Review of Pertinent
Literature.)
y.
Figure l4. Equilibria in pyridoxal - valine or pyrldoxalN-methochloride - valine solutions (Johnston
, 1963)
124
K,
PV=
P" + V
'
- H+
not present in pyridoxal
methochloride imines
K."
p- + VPV- H+
(log K," = 1.52,PLMj
3.47,PL)
PK,2pv
+
+
p- + V
8.0 ,PLK
5.9, PL
PV
(log K, ' = -0.02 ± 0.1,PLMj
^
' = K-î.
-0.27,PL)
K,2p
^pv • ^3pv
Kl" = %• ^
Spv
- K^".
Kgpv
125
Equilibrium Constants for the Formation of Aldimines
of Leucine and Deoxypyridoxal, Leucine and Pyridoxal
Phosphate as a Function of pH
(Most of the experimental work, calculations and curve-fitting
was done by Erika Rommel, Louise Hodgin, and David E. Metzler,
Ames, Iowa, Iowa State University of Science and Technology,
Department of Biochemistry and Biophysics. Private communi­
cation. 1964)
The appearance of absorption bands in the region 400 to
420 mp. indicated the formation of imines in solutions of pyri­
doxal phosphate and leucine and of deoxypyridoxal and leucine.
The extent of formation of these imines is dependent upon the
amino acid concentration (or the ratio of concentrations of
amino acid and pyridoxal analog, since imine formation is an
equilibrium) and on the hydrogen ion concentration of the
solution. By measuring the relative extent of imine formation
spectrophotometrically in solutions of' varying amino acid
concentration, at constant deoxypyridoxal (or pyridoxal
phosphate) concentration and constant pH, an imine formation
constant was calculated by the graphical method previously
described for imine formation from pyridoxal methochloride
and Valine (Experimental).
The equilibrium constants for the formation of these
imines were then determined as a function of pH, and the
logarithm of K
values (observed imine formation constants)
were plotted against pH (Figure 15). The shape of the curve
for imine formation from deoxypyridoxal and leucine was similar
Figure 15. Variation of the logarithm of the equilibrium constants for imine
formation with pH
Imines formed from:
O
Deoxypyridoxal + leucine (K^)
• Pyridoxal phosphate -h leucine (K )
o
A Pyridoxal anion + valine anion (K^)
T
T
Log KpH
1
M
ro
0
-2
3
4
5
128
to that from pyridoxal and valine, which was reported by
Metzler (1957). Experimental points were in agreement with
I
the theoretical curve calculated by a method previously de­
scribed (Metzler, 1957) and outlined In Figure l6, using the
pKg^ values of the species involved and the Kj^' value for the
hydrogen-bonded imine.
Prom the calculated formation constants the molar frac­
tions of total imine formed from 0.01 M deoxypyridoxal, 0.05
M deoxypyridoxal, and 0.05 M leucine were calculated by a
method of successive approximations (instead of solving
quadratic equations), and plotted against pH (Figure 22).
The molar fraction of protonated imine (protonated at the
pyridine ring nitrogen atom) was calculated from equation
(17) relating the fraction of total imine at each pH value,
assuming pK
= 6.10 (Figure 16). Evidence for this pK
2P—L
was obtained from the intersection of curves for a plot of
the molar absorbancy Indices of the imine at 4l4 and 294 mji
(which in turn were obtained from intercepts found by extra­
polation in the plots used to determine the imine formation
constants) as a function of pH.
(17)
(Fraction total imine) (l-f
) = (Fraction imine H)
P-L
A few experimental points for the observed equilibrium
constants for the formation of imlne from pyridoxal phosphate
and leucine were included in Figure 15.
The shape of the
curve (for log KQ VS. pH) was assumed to be similar to those
for imlne formation from deoxypyridoxal and leucine and from
Figure 16. Equilibria of imine formation from pyridoxal
phosphate (PLP) and leucine, deoxypyridoxal
(DPL; and leucine, and equations for calcula­
tion of formation constants
130
P-L=
pK
= 10.5,PLP
3P-L 11, DPL
K,
+
pK.
2P
P±
pK.
IP
8.69,
8.14,
P-L
PLP
DPL
pKg^ =9.60
4.14, PLP
4.17, DPL
pKiL = 2.36
PK
2P-L
+
6.1 DPL
P-LK."
P+
+
T +
P-L"*
J. ^ (P-L) _ ^ (P-L+) + (P-L*) + (P-L-) + (P-L-)
o (P (L)
(P++ p± + p-)(L+ + L± + L")
_ (P-L-)
1 , (H+) . ^2P-L , ^2P-L^3P-L
(pi)
(H+)
(H+)2
K,
IP-L
(' *
(' *
H"
4-
--K^.
KIP-L
H+
(•1+ K
-h
IP
log K. - log K.' = lo
- log
•
(H+)
1+
KlP
l + ifj.
K.IP-L
KGP-
*2P-L
(H"^)
(H+)
+-^ - log
(H+)
L
KlL
^P-L%P-L
(H+)2
%
2L
(H+)
I
131
pyrldoxal and valine. Since the curves are parallel, the
relative extent of imine formation from these pyridoxal analogs
is also dependent on pH.
/
Imine concentrations were calculated from direct spectral
data and compared to values predicted for the corresponding
conditions (Table j):
Table 7. Comparison of imine concentrations from direct spec­
tral data of reaction solutions at time zero with
predicted from the theoretical curve (Figure 15)
0.10 M leucine + M DPL,
pH
(DPL-leu imine), M X 10^
specjtral
predicted
0.00113
4.6
6.3
6.02
0.01 i
4.2
29 ± 1
43.7
0.01
7.1
48
67.5
The graphical methods of Harbury and Foley, (1958) and
of Isenberg and Szent-Gyorgi (1958) also could have been
used
for the evaluation of imine formation constants from the data
obtained.
These methods may have some advantage over that of
Ketelaar et al., (1951, 1952), in that the apparent molar
absorbancy indices for solutions of imine do not have to be
calculated, since absorbancies are used directly, and the molar
absorbancy index and concentration of pyridoxal analog would
not have to be determined. (Compare equations given in Review
132
of Pertinent Literature.)
Nonenzymio Transamination of Leucine with Deoxypyridoxal
The roles of the nitrogen atom of the pyridine ring, the
3-phenolio, 4-formyl and 5-hydroxymethyl groups of pyridoxal
(vitamin Bg aldehyde), the most thoroughly studied amino group
acceptor in nonenzymic transamination of amino acids, have
been discussed (Review of Pertinent Literature).
However, the
effects of these groups and of other substituents at the 5position of pyridoxal (Figure 17) have not been studied quan­
titatively in such model systems.
In order to quantitatively
measure the effects of these groups in a model transamination
system, 5-deoxypyridoxal, which has a methyl group in place of
the hydroxymethyl group of pyridoxal, was studied as a refer­
ence amino group acceptor.
Leucine was chosen as the amino group donor because of
its relatively high imine formation constant (Metzler, 1957)
and transamination rate, (Metzler and Snell, 1952b) compared
to that of other naturally-occurring amino acids, and^its
expected lack of side reactions encountered with amino acids
(such as serine, threonine, glutamate, cycteine, lysine), having
extra functional groups. Furthermore, its transamination
product, alpha-ketoisocaproate was expected to be more stable
than keto acids with smaller side chains, such as pyruvate
(from alanine).
The temperature of 25°C was chosen for this kinetic model
Figure 17.
Model systems of nonenzymic transamination between pyrldoxal analogs
and amino acids (See Figure 1 for nomenclature of analogs)
©
H^COe®
I3H
f
or
'H.
-H2O
® Y
•\©
(PL T=1NF1L06)
RLDIMIKIE
/
^(KIC)
/
+ H30®
©
KETIMINE
(pM -RKIFILOG)
5-substituted PL analogs (intramolecular catalysts)
Z = -OH
PL
Pyrldoxal; PM Pyrldoxamine Analogs
- -OPO K3
- -H
PLP
Pyridoxal-5-phosphate
DPL
5-Deoxypyrldoxal
- -0_"
CPL
"5-Carboxypyridoxal" or "5Pyridoxalylate"
= -COg-
PPL or PLP
"alpha^-Pyridoxalylformate"
= -CHgCOg-
APL or PLA
"alpha^-Pyridoxalylacetate"
R = -CHgCHfCHg)^
for leucine
^
KIC = alpha-ketoisocaproate
B = general base
catalyst (intermolecular),
e.g., buffer
134
system, because formation and acid dissociation constants of
Imines of pyridoxal and amino acids are known, or can readily
be determined, and pH measurements are more reliable at room
temperature than at 100°C.
This information is necessary for
the calculation of precise rate constants, because the true
reaction rate is a measure, of not the amount of pyridine
aldehyde lost or of keto acid formed, but of the amount of
transamination intermediate converted to products in a unit of
time.
This nonenzymlc transamination does proceed at 25°C, and
may be followed both qualitatively and quantitatively by
several techniques.
Identification of products has been made
by paper and thin layer chromatography.
The rates of change
in pyridoxal analog and in keto acid have been followed by
several chemical- spectrophotometrlc methods.
Spectral changes
in a reaction solution (attributed to changes in concentrations
of reactants. Intermediates and products) are in quantitative
agreement with rates obtained by the chemical methods.
Chromatography of reaction solutions
Ketoisocaproate (KIC) was identified as its 2,4-dinitrophenylhydrazone (DNP), most conveniently, by separating this
yellow DNP-derivative from that of deoxypyrldoxal on a thinlayer silica gel plate.
This was qualitative evidence for the
production of KIC from the nonenzymlc transamination of leucine
with deoxypyrldoxal.
Usually two yellow spots were detected
135
for KIC, corresponding to the syn- and anti-forms of DNP-KIC
(Metzler and Snell, 1952a). A yellow spot for the DNP-pyridoxal analog had a slightly lower R value than those of DNPF
KIC. The DNPH reagent itself lagged a little behind the
solvent front and appeared as a pair of yellow spots or as a
streak. (See Table 8 for comparison of R values).
F
Constituents with amino groups were identified with a
ninhydrin spray reagent.
Leucine, having a low R^ value of
0.3J was easily identified as a purple or pink spot. Deoxypyridoxamine (if present in high enough concentration) gave an
orange spot which migrated very little from the origin, as did
a pyridoxamine standard (Figure l8).
No para-unsubstituted phenols other than deoxypyridoxal
and deoxypyridoxamine, and perhaps the leucine aldimine of
deoxypyridoxal were detected by Gibbs' reagent. Thus, no
intermediates or products ; other than those expected from the
nonenzymic transamination of leucine with deoxypyridoxal to
yield ketoisocaproate and deoxypyridoxamine were detected by
these chromatographic techniques. These results are further
evidence that the changes observed by quantitative analytical
techniques, especially spectral, are indeed related to an
actual nonenzymic transamination process.
Stoichiometric and kinetic anomalies
By comparing the changes in concentrations of ketoiso­
caproate and deoxypyridoxal in the reactions studied at dif-
136
Table 8. Chromatography of a solution after about 200 hours
reaction, 25°C
Initial conditions:
50 mM leu, 10 mM DPL, pH 6.6, unbuffered
dlnitrophenylhydrazine (DNPH) solutions spotted; ninhydrin
Nin) solution sprayed after development)
R„ values on paper - developed with upper layer of n-BuOH (5),
HgO (4), EtOH (1); BEW, 514
DNP - KIC
0.86
yellow
Nin - leu
0.59
purple
DNPH
0.95
yellow
lit. values
0.79 - 0.bb&
R™ values on thin-layer silica gel plates - developed with
upper layer of n-BuOH (5), HgO (4), EtOH (l)
Standards
DNPH
( 2 ) 0.67 - 0 . 7 6
yellow
DNP - KIC
DNP - DPL
0.59
0.50 - 0.53
0.46 - 0.49
yellow
yellow
yellow
0.54 - 0 . 6 2
0.45 - 0.51
0.62 - 0.68
Nin - leu
0.35
purple
0 . 3 1 - 0.4
Nin - DPM
0.05 - 0 . 0 9
orange
0.05 - 0.09
(Nin - PM)
Developed with n-BuOH (4), HOAc (l), H^O (5)
DNP - KIC
0.7
yellow
Nin - leu
0.3
purple
Nin - DPM
0.06
orange - red
R|. values of spots from Gibbs' Reagent sprayed on thin-layer
chromatograms not treated with DNPH; developed with upper
layer on n-BuOH (5), H^O (4), EtOH (l)
DPL
0.65
DPL - leu
0.31
aldimine
aSource:
Meister and Abendschein, 1956
Figure 1 8 .
Identification of amino group acceptors and
donors in nonenzymic transamination by thinlayer chromatography
Standards and samples
conc, M
no. spots
Reagents
spot
no
1 PM
0.01
15
2 Leu
0.16
5
3 DPL
0.001
25
DNPH
8
4 DPL + Leu
50
(8 X 10-4 + 0.03, pH 6.2)
DNPH
12
5 4 at tg
50
DNPH
12
20
DNPH
12
7 as 4
50
Gibbs'
20
8 as 5
50
Gibbs'
20;
40
Gibbs'
20
6 KIC
0.05
9 DPL
0.001
10 PM
0.01
Solvent system:
Results:
1
2
3
4
5
6
T
8
9
10
Ninhydrin
spray
after
develop­
ment
Gibbs'
20
15
upper layer of BEV, 514
Color (identity)
Photo taken 1 day after
color development
Orange (PM - Nin)
Purple (Leu - Nin), (color fades almost completely
in one day)
Yellow (DPL - DNP)
DNPH (2)
Purple (Leu - Nin); Yellow (DPL - DNP)
Orange (PM - Nin);
Purple (Leu - Nin); Yellow (DPL - DNP); (KIC - DNP)
(2)
Yellow (KIC - DNP) (2)
Purple (Leu - Nin);
Brown, Gray
(DPL - Gibbs')
Orange (PM - Nin);
Purple (Leu - Nin); Yellow (KIC - DNP) (2); Brown
Brown (DPL - Gibbs')
(DPL - Gibbs')
Orange (PM - Nin);
Brown (DPL - Gibbs')
138
origin
solvent front
139
ferent pH's (Table 9)j It can be noted that initially,
je.,
after 48 hours reaction, the changes in concentration of deoxypyridoxal are at least equal to, but usually greater than
(depending upon pH) the corresponding changes in concentration
of keto acid.
These concentrations were calculated from
standard calibration curves based on loiown amounts of keto
acid and pyridoxal analog (See Experimental, Figure 8). Howeve^, at a much later time in each of these reactions (ap­
proaching equilibrium), the changes in concentration of
deoxypyridoxal are less than the corresponding changes in
concentration of keto acid, except at the pH 4 optimum (Figure
19). This greater initial loss and lesser later loss of deoxy­
pyridoxal than would be expected from the corresponding keto
acid formation can not be explained solely on the basis of a
side reaction Involving transamination products.
It has generally been observed that the rate of decrease
in pyridoxal analog, followed by decrease In absorbancy at 400
mp by the qulnolylhydrazone (QH) method, is measurably greater
than (up to 3 times greater under certain conditions of pH and
concentration, especially at pH 11.6) the corresponding rate
of Increase in ketoisocaproate, followed by increase in absorb­
ancy at 305 mp by the Q.H method (Table 9).
If it is assumed
that the qulnolylhydrazone method is able to detect aldimine
and ketimlne intermediates, as well as pyridine aldehyde and
alpha-keto acid, then overall rates of the transamination
reaction would be the same, regardless if they were based on
i4o
Table 9. Stoichiometric and kinetic results of transamination
of leucine with deoxypyridoxal®-
Stoichiometric
(changes in millimolar
concentrations)
Kinetic^
kobs ^ 107 sec:'
KIC
DPL
KIC
DPL
1.2
0
0
1°
0
1.7
4.1
0
0
1®
0
504
0.9
2. 1
4.5 • 8.4
5.75
648
1.0
1.35
6.9
pH after after
48 hrs.
t hrs.
KIC
DPL
11
(16)
6.15
_5.2
17.5
• ro
00
1.1
768
0.65
0.9
8.0
5.6
4 (8)
±_2
648
1.0
1.0
7.7
6.4
9.35
648
1.3
1.7
7.7
6.4
11.65
456
1.8
3.4
7.1
5.9
12.55
456
1.5
1.75
7.3
6.8
^^8
^Initial conditions:
25°C
4.5
iL2.5
6.4
6.1
(11)
6.5
(24) 15
:t.5
37
+9
^2.5
15
±.2
13
J.3
17
^10
50 mM leu, 10 mM DPI, unbuffered,
^Calculated from adjacent Guggenheim points and/or de­
termined from Guggenheim plots; pH profile - Figure 21
(Initial rate -- deviation from first order plot during
at least first 24 hrs.)
•^(Estimated from approximate ti as compared to reactions
at higher pH, during first ^ week of reaction, but
little further reaction detected after two months
Figure 19.
Rate curves for reaction between leucine and deoxypyridoxal:
Absorbancy vs. time (hours), indicating (itiM DPL or KIC at
0; 48 J 504 hours).
Initial conditions:
0.05 M leu, 0.01 M DPL, unbuffered, 25°
Aliquots analyzed by QH method
OA,
mu = f (DPL)
400 '
A A205
pH 4.1
pH 1-7
6
= f
(KIC)
(DPL)
(KIC)
^(L—ùS—à
100
200
300
TIME
400
(HRS)
500
600
1000
143
absorbancy at 305 or at 400
This assumption seems to be
valid because rates based on other analytical methods, includ­
ing the ethanol imine method and direct spectral measurements
used in other experiments, are in agreement with those obtained
from the QH method.
Side reactions involving the pyridoxal analog are sug­
gested to account for this anomaly. Such reactions may include
air oxidation of the pyridoxal analog or formation of the imine
from the pyridoxal analog and its corresponding pyridoxamine
analog, as suggested by Banks £t
(1961) and by Bruice and
Topping (1963a). In the related study by Banks £t ^., (1961)
the straight line curves obtained from the reciprocal plot of
the data fitted the equation derived for spectrophotometric
determination of the imine formation constant (See Review of
Pertinent Literature). This indicated the formation of
ketlmine from pyridoxamine and pyruvate at pH above 8 involved
equimolar amounts of reactants.
At pH 10, however, the forv/ard reaction between pyridoxal
and alanine did not yield equimolar amounts of products, but a
2 to 1 ratio of pyridoxamine to pyruvate. This stoichiometric
anomaly was explained by possible side reactions such as air
oxidation of pyridoxal or decarboxylation of pyruvate. Fur­
thermore, a reaction between pyridoxal and pyridoxamine at pH
11 was found to yield twice the initial molar amount of pyri­
doxal. This reaction was explained on the basis of air oxi­
dation of an imine intermediate (Banks, et al., 1961).
144
In another kinetic study of nonenzymic transamination, a
greater rate of cleavage of the ketimine of phenylglycine and
pyridoxal in alcohol, as compared to aqueous solution, resulted
in stoichiometric amounts of products and indicated negligible
side reactions (Bruice and Topping, 1963a). However, this
imidazole-catalyzed reaction in aqueous solution (pH 8.6) had
side reactions resulting from the slow breakdown of ketimine.
Imidazole catalysis had little effect on ketimine cleavage,
although it facilitated the rate-determining aldimine to
ketimine transformation.
At less than 0.3 M imidazole the
formation of imine from pyridoxal and pyridoxamine was sig­
nificant and rate constants could be determined (Bruice and
Topping, 1963d).
In these experiments air oxidation of pyri­
doxal was minimized by several precautions.
Side reactions of an uncharacterized nature occurred also
in alkaline medium (pH 9.6) in the oxidative deamination of
amino acids by pyridoxal (Ikawa and Snell, 1954b). There was
a greater loss of pyridoxal than expected, since pyridoxal
decreased in smaller amounts than the ammonia formed. In
some reaction mixtures 4-pyridoxlc acid, the oxidation product
of pyridoxal with the 4-formyl group converted to a carboxylic
acid, was found.
The rate of racemization of alanine in the reaction with
pyridoxal, observed by changes in optical rotation, did not
equal the rate associated with the third relaxation time in
spectral measurements (Fleck and Alberty, 1962). Both meas­
145
urements were expected to agree since both depend on the
aldimine-ketimine tautomerization (in which the opticallyactive center at the alpha-carbon is lost after removal of
the proton in the rate-limiting step in transamination).
Some of these stoichiometric and kinetic anomalies may
be due to the instability of aqueous solutions of pyridoxal
observed by Christensen (1958), Davis _et al., {19S1), Fleck
and Alberty (1962). Spectral changes were observed even in
sterilized solutions of pyridoxal at pH 8 kept in total dark­
ness over periods of days at room temperature (Fleck and
Alberty, 1962). The total magnitude of these absorbancy
changes per mole of pyridoxal was about 50 of the changes
observed when alanine was also present at concentrations of
about 0.2 M, but the magnitudes of these changes were not
reproducible.
However, deoxypyridoxal, 0.01 M in an un­
buffered solution of 0.05 M leucine at pH less than 2, was
stable for at least 1000 hours, since there was no signifi­
cant change in absorbancy of aliquots analyzed by the quinolyihydrazone method at 400 mp. (Figure 19).
Direct spectral measurements were obtained for two
reactions in unbuffered solution near neutral pH. For the
reaction with an initial pH of 6.6, decreases in absorbancies
at 415 and 234 mp, maxima for aldimine absorption followed
pseudo first-order kinetics. The rate constants were 7.I and
10 X lO'^/sec, respectively, as compared to those obtained
from data by the QH method: 6.9+0.8 and 5.9+0.5 X lO'^/sec,
146
for increase in absorbancy at 305 mp. and decrease in absorbancy
at 400 mju, respectively.
An example of such spectral changes may be seen in Figure
20.
In this reaction of deoxypyridoxal with leucine the
maximum absorbancy at 288 np increased and was also followed
as a function of time. For the reaction with an initial pH
of 7.1 J decrease in aldimine absorbancy at 420 mp and increase
in absorbancies at 325 and 285 mp. (probably due to ketimine
and/or deoxypyridoxamine products) also yielded pseudo firstorder rate constants, but that obtained from the data at 325
mp was about twice the rates obtained from data at 420 and
285 mp.
All of these rates, however, were considerably higher
than the corresponding constants from QH data (Table 10).
Table 10. Comparison of rate constants from the quinolylhydrazone analyses with those from direct spectral
analyses of the reaction of leucine (leu) with
deoxypyridoxal (DPL)&
Direct spectra
QH
1
305
0
0
k
X lo'^sec'^:
obs
+A
<
Analytical
method:
7.2
11.9
^Initial conditions:
unbuffered, 25°
"^420 *^325 "*"^285
15.6
34.7
19.6
0.01 M Leu, 0.01 M DPL, pH 7.1,
147
Spectral changes noted in the deoxypyridoxal control
solution may be due to photodecomposition or air oxidation
(Figure 20). Furthermore, spectra and ethanolimine analyses
of deoxypyridoxal (DPL) solutions showed no significant
changes during the first day after mixing.
However, gradual
pH and spectral changes occurred especially in unbuffered
solutions of DPL after one day, and indicated loss of the
free aldehyde form at about 380 m^ and Increase in the 325
mp. absorption band (See Figure 20). Air oxidation may be at
least partly responsible since these changes were inhibited
by occassionally bubbling nitrogen gas into the solutions
(Robert Johnson, Ames, Iowa, Iowa State University of
Science and Technology, Department of Biochemistry and Bio­
physics.
Non-enzymatic transamination reactions of amino
acid esters with pyridoxal analogs.
Private communication.
1964.)
Variation of pH in unbuffered reaction solutions was
considered as a possible cause of the kinetic anomalies. The
pH values used in Figure 21, the pH profile for the reaction
rate of leucine and deoxypyridoxal, were obtained during the
first day of reaction. These reactions were followed over
periods of from three to eight weeks, and pH changes were
noticed in these unbuffered solutions.
These changes varied
both in magnitude and in direction from the initially deter­
mined pH values.
Changes of 0.5 to 1 pH unit were recorded
after about four weeks reaction. These changes are indicated
Figure 20, Spectral-tljne studies at 25°C of
a) an aqueous solution of DPL (unbuffered)
spectrum no.
3^
T
4^
5
time after dilution
.
55'
6 hr. 25'
21 hr. 25'
48 hr.
pH
5.00
6.8
b) an aqueous solution of DPL in 0.10 M
Leu (unbuffered)
1
1<
2
3
4
5
35'
5 hr. 10'
21 hr.
5'
47 hr. 50'
4.64
6.6
Spectral analysis of DPL in O.IN HCl:
1.13 X 10 ^ M DPL in a) and b, 9.0 mm
+0.01
cell spacers required
^Spectra not shown
149
1
1
i
1
1
1
1
1
1
1
1
1
1
1
412
-- b)DPL
+Leu
pl-Â
/ A
q) DPL
37T
1
A~"'
<325
5-j\
/L r\
/M
//
/ V/
/
1-^
1
1
\
1
1
1
1
1 1
300 350 400
^50
300
WAVELENGTH ^ m;u
350
400
-
150
by the sizes of figures (width of triangles or rectangles)
surrounding the experimental values of Figure 21. (Also, see
Figure 20 for pH changes observed in unbuffered solutions of
deoxypyridoxal).
It was thought that the pH of an unbuffered solution of
leucine and deoxypyridoxal would remain fairly constant through­
out the reaction, since there is no net gain or loss of protons
in the nonenzymic transamination reaction (Figure 17). Also,
the pH values of unbuffered solutions of pyridoxal phosphate
and leucine remained fairly constant over periods of several
days.
Although boiled, redistilled water was used in the pre­
paration of reaction solutions, pH changes observed in these
solutions might be due to reabsorption of carbon dioxide and
oxygen from the atmosphere when the tubes containing the
solutions were opened to withdraw aliquots for analysis.
In some cases the Guggenheim plots used to calculate the
pseudo first-order rate constants, based on the deoxypyridoxal
data were concave downward and initial and final rates could
be calculated from two linear sets of points (Figure 10).
Fleck and Alberty (1962) also noted a curvature concave down­
ward in' some of their Guggenheim plots during the first 10^ of
the time period. These deviations seemed to occur predomi­
nantly at low pH values or at high amino acid concentrations.
Perhaps, intermolecular acid-base catalysis of two molecules
of imine or by amino acid anion with one molecule of imine
would explain this departure from pseudo first-order kinetics.
151
Another possible side reaction to account for greater loss
of deoxypyridoxal is the condensation of the alpha-carbon of an
imine with the aldehyde group of deoxypyridoxal. This con­
densation was shown to take place in the reaction of pyridoxal
phosphate with cycloserine (Roze, 1964).
In considering these kinetic and stoichiometric anomalies,
it seems that the most reliable rate data are probably those
based on keto acid formation, or on loss of aldimine followed
spectrophotometrically, until further investigation into the
nature of the possible side reactions of pyridoxal analogs
yields an explanation which quantitatively accounts for these
anomalies.
Rate of reaction as a function of pH
Pseudo first-order rate constants for transamination of
deoxypyridoxal by a 5 to 1 excess of leucine, in unbuffered
aqueous solutions, were determined and plotted against the
initial pH of the reaction solutions in Figure 21.
Two rate
maxima exist - one at about pH 4 and a higher one above pH 11,
with a minimum between them at neutral pH values. It was
observed that reaction solutions of pH less than 3 had only a
slightly yellowish tinge compared to brighter and deeper
yellow solutions at higher pH, Indicating relative extent of
Imine formation as a function of pH in these solutions. The
bright yellow solutions also faded relatively rapidly corre­
sponding to their rates of conversion of imine to products.
152
In order to Interpret this pH profile for the rate of nonenzymic transamination quantitatively, the pH profile for the
corresponding imine formation constants must be used (Figure
15).
Consider the reaction, P + L
. P-L ^i s K
PM,
TSFt
slow
where P = pyridoxal analog, PM = pyridoxamine analog, L =
leucine, P-L = imine intermediate, K = keto acid (from leu),
and equilibrium constant for imine formation (observed at any
(P—L)
PH), KQ =
&nd k^ is the first-order rate constant for
conversion of P-L to products. First-order rate constants for
formation of keto acid in unbuffered solutions may be expressed
by equation ( 1 8 ) .
(18) k
= -(a(r-L/at)o ^ 4(d(K)/citlb
^
(P-L)
(P-L)
The imines, P-L, are formed rapidly and the rate measured for
loss of pyridoxal analog, formation of keto acid, or loss of
aldimine, by several methods, must be proportional to that of
the subsequent conversion of the imine to products.
The optimum at pH 4 may be consistent with the reactive
Intermediate being the hydrogen-bonded aldimine with a protonated nitrogen atom in the pyridine ring.
This protonated
imine species becomes predominant below pH 6 (pK 6 .I). Below
pH 4 the observed rate constant falls off sharply, probably
due to the decrease in the extent of imine formation. In this
low pH range the cation forms of deoxypyridoxal and an amino
acid would not be expected to form a very stable imine (a
dication). Furthermore, dehydration of the carbinolamine
Figure 21.
Pseudo first-order rate constants for trans­
amination of leucine with deoxypyrldoxal as a
function of pH
i
Initial conditions: 0.05 M Leu, 0.01 M DPL, unbuffered,
250
QH analytical method
Ak
OBs
for KIC formation from
305 mu
for DPL decrease AA , _
400 mu
©estimated values (see footnote a in Table 9)
Qk ,
.OBs
Guggenheim kinetic method used where k^^^ - -2.303
log AA/At, equation (16), calculated from pairs
of adjacent points to obtain average; sizes of
symbols indicate relative certainty in k^^^
(average error) and in pH of solutions over time
of measurements; Table 9 data
pseudo first-order rates of re­
actions having deviations from linear Guggenheim
plot: . initial rates from measurements during at
least first 24 hours and 24 + T hours of reaction,
where T is > t|- (See Experimental for sample
plots obtained: Figure 10)
'^obs^'O^ sec
ro
O
X
—T—
Ol
—T"
VJl
i
1
155
Intermediate in semlcarbazone formation from pyridoxal has been
found to be rate-limiting below pH 4 (Cordes and Jencks, 1962a,
b).
However, at about pH 4 the dipolar ions of deoxypyridoxal
and amino acid form a more stable imine, which also readily
breaks down to transamination products, because the positivelycharged pyridine ring nitrogen atom aids in withdrawal of
electrons from the alpha-carbon atom of the amino acid residue.
As the pH increases toward neutrality the reaction rate
decreases, probably as a result of dissociation of the pyridinium group (pK 6), and thus decreasing concentration of the
reactive protonated imine, and formation of increasing concen­
tration of the relatively unreactive imine having no positive
charge on its pyridine ring nitrogen atom. The fractions of
total, (l)/(P + I), and protonated, (IH'*')/(P + l), aldimine
from leucine and deoxypyridoxal were calculated from the imine
formation constants as functions of pH and from pK^p ^ of 6.1,
where I is P-L. These values are presented in Figure 22 along
with rate constants obtained for the formation of ketoisocaproate in the nonenzymic transamination, pH 4 to 10.
Keto acid may be formed primarily from the protonated
aldimine below pH 7 and the rate optimum in the acid region
may be explained on the basis of a direct proportionality
relationship of the rate to the fraction of concentration of
protonated aldimine:
I •+
observed rate = k(IH*).
An alternative explanation for the rate of the acid pH region
is that it may be proportional to the product of total imine
Figure 22. Fraction of deoxypyridoxal-leucine aldiraine as
a function of pH
Upper curve; Fraction of deoxypyridoxal as aldimine calcu­
lated from imlne formation constants, by a
method described earlier (Metzler, 1957), out­
lined in Figure l6, and from successively
approximating the imlne concentration in
solutions of 0.01 M DPL and 0.05 M leu, at
each pH
Lower curve: Fraction of protonated imine calculated from
pK
=6.1 (verified by a plot of a„
vs.
P-XJ
pH)
A Transamination rates: (Figure 21, Tablé 9)
V.O
—
FRACTION OF IMINE FORMED
>0
vO
O
V.O
n
O
G
a
T
O
l
O
:o
Oo
'<obs.^l0^sec'
o\
o
Oo
-
jO
S
V.O
C
o
C
vO
Û
158
concentration and hydrogen Ion concentration at each pH: rate
= k(H+)(l).
Above pH 7 the rate begins to Increase, despite the de­
crease in the fraction of protonated imine below 10^.
Although
the fraction of total or unprotonated imine is greater than
90% from about pH 7*5 to above 10, the increase in rate in the
alkaline pH range may be partly due to general base catalysis
of the conversion of aldlmine to ketimine by the excess amino
acid anion, by another molecule of unprotonated aldlmine acting
Intermolecularly or by hydroxide ion, in the extraction of the
hydrogen from the alpha-carbon of the amino acid residue in the
unprotonated imine.
However, at very high pH, imine formation
decreases probably as a result of dissociation of the imine
hydrogen-bonded proton (pK^ 11). Hydrogen-bonding probably
stabilizes the imine and loss of this stabilizing effect on
imine formation probably begins to show up in the over-all
transamination rate at least at pH 12.
The pH profile for transamination of leucine by deoxypyrldoxal, may be compared with that obtained for transamina­
tion of alanine by pyrldoxal (Blake £t aJ., 1963).
At 100°C,
in the absence of added buffer the latter had a broad maximum
near pH 5 and a much higher, sharper peak at pH 9.5 (pH's
measured at 25°).
Other pyrldoxal model transamination re­
actions studied Indicated that the rates Increased from pH 7
to 10 in buffered systems with phenylglyclne (Bruice and Topping,
1963b) and with alanine (Banks et al., 196I; Fleck and Alberty,
159
1962).
Paster nonenzymic transamination rates were observed at
acid pH values than in neutral or slightly alkaline solutions
(Banks et al., 1961).
Similar results were obtained from re­
actions of pyridoxal phosphate and amino acids catalyzed by
metal ions (Cattaneo £t a^., 196O; Review of Pertinent Liter­
ature).
The optimum pH for metal ion-catalyzed transamination of
glutamate with pyridoxal depended somewhat on the metal ion,
according to Longenecker and Snell (1957). For the best metal
ion catalysts, Cu (ll), A1 (III), Pe (II, III), the optimum
pH was about 4.8, but for Zn (II) and others, it was near pH
7.
An optimum of about pH 5 was found for unbuffered solutions.
In general, metal ion-catalyzed transamination of amino acids
with pyridoxal had a maximum rate near pH 4.5 in the range of
pH 2 to 9 at 100°C and at about pH 5.2 at 25°C (Metzler and
Snell, 1952b).
Little or no transamination was found to occur
at pH 9.6, where certain amino acids are oxidized optimally.
The question arose as to whether the pH 4 optimum was due
to traces of metal ion in these unbuffered systems.
Reaction
solutions containing 1 mM EDTA (ethylenediaminetetraacetate),
a metal ion-chelating agent, were compared with those lacking
EDTA under the same conditions.
No effect of inhibition of
catalysis by metal ions was observed, within experimental
error in determination of the rate constants (Table ll).
However, acetate buffer at a concentration of 1 M, may
have complexed with any contaminating metal ions and thus no
effect of EDTA could be observed.
Indeed, in reaction solu-
l6o
Table 11.
Observed pseudo first-order rate constants (X 10^/
sec.) for nonenzyniic transamination®-
QH method
^305
No EDTA
"\oo
2.8
+1.0 mM EDTA 2.60
+0.3
(DPL)
Direct spectra
'^413
(aldimine)
2.8
3.0
3.1
3.3
^Initial conditions: 0.10 M leu, 0.01 M DPL, 1.0 M
acetate, pH 4.2, 25° (Although Guggenheim plots are not
linear over the time period studied the slopes were taken
as the best straight lines through most of the points).
tions which had the acetate buffer concentration decreased by
a factor of 100, a small effect of EDTA was noted.
Although
quantitative rate constants were not obtained from these QH data,
the relative rates found by taking the inverse ratio of times
to attain the same amount of reaction, as measured by changes
in absorbancies at 305 and 400 m^ in QH-treated aliquots,
indicated that the reaction in the absence of EDTA was at least
about 1.5 times faster than the reaction in its presence (but
still only half as large as the effects observed by Metzler
and Snell, 1952b). Also, the yellow color (due to aldimine)
in the reaction solution containing no EDTA faded faster than
that containing EDTA.
However, it is not expected that this
l6l
small rate enhancement by contaminating metal Ions in unbuff­
ered systems would significantly shift or eliminate this acid
pH optimum, especially since Cennamo (1964) found that the
addition of metal ions does not enhance the transamination of
amino acid esters by pyridoxal at 100° - a reaction which has
an optimum at pH 5.3, measured at 25°.
Effect of leucine concentration
The rate of KIC formation at pH 7.1 to 7.2 increased by
a factor of more than two (Table 12) when the leucine concen­
tration was doubled (from 5/1 to a lO/l excess over that of
deoxypyrldoxal).
This may be evidence of general acid cata­
lysis by leucine.
The observed overall rate constants were
changed to k'^ values by dividing by the appropriate leucine
concentration, equation (19). For pseudo first-order con­
ditions (excess leucine). -d(P)/dt = ^obs^^^ ~ k^Q(L)(P),
where
is the observed rate constant, and k^ is the pseudo
first-order rate constant in an unbuffered system.
(19) k'c =
In 0.50 M acetate buffer solutions near the pH optimum,
leucine concentration was varied between 0.01 and 0.1 M with
constant deoxypyrldoxal concentration at 1.0 mM.
Pseudo first-
order rate constants were determined from Guggenheim plots of
logarithms of absorbancy changes (increasing at 325 mp and de­
creasing at 4l5 and 400 mp) vs. time.
Ketolsocaproate and
deoxypyridoxamlne were Identified as reaction products by the
l62
Table 12. Effect of leucine concentration on pseudo firstorder rate constants^
k
QH method: +
mp (KIC)
X lO^/sec:
2.65
v,
7.2
k' X loVsec:
o
mji (DPL)
7.9
("Initial" rate In
Guggenheim plot)
2.35 ("final" rate In
Guggenheim plot)
53v
Y2°
^Initial conditions: 0.05 M Leu, 0.01 M DPL, pH 7.2, 25°
^Same conditions except 0.10 M Leu
thin-layer chromatographic techniques. (These experiments
were carried out by Louise Hodgin in this laboratory.) The
linear relationship of the observed rate constants with in­
creasing leucine concentration (Figure 23) may be partly due
to increasing concentrations of reactive Intermediate at this
pH, where aldimlne formation constants are not optimal (Figure
15).
The fraction of deoxypyridoxal calculated as Imine present
under these conditions increases non-llnearly from 0.44 at 0.01
M leucine to 0.88 at 0.10 M leucine.
Such a curve would
intersect, of course, with the origin (no leucine), and coin­
cide with the first three points of Figure 23. Thus, this
Figure 23. Effect of leucine concentration on the pseudo
first-order rate constant for the nonenzymic
transamination of leucine with deoxypyridoxal
6
P
X 10 /sec.plotted vs. (leu), M X 10 , where k
o d b
obs
values are averages of three determinations, with lines
extending from points indicating average deviations, from
Guggenheim plots of logarithms of absorbancy changes
Aoo'"^^325 V)
time-
a
Initial conditions: 1.0 mM DPI, 0.50 M acetate
buffer
(leu), M
0.01
0.02
0.05
0.10
pH
4.3-4.4
4.2
4.1
4.05
164
^ 06
04
02
2
4
6
[Leu]
M * 10^
ô
10
12
165
part of the rate-leucine dependence curve (Figure 23) can be
explained on the basis of increasing formation of imine, but
at 0.1 M leucine the greater rate observed than would be
éxpected from this relationship could have been acid-base
catalysis by leucine dipolar ion.
From Figure 23 the parameters in equation (l9a) may be
determined.
(19a) k '
= k' (L) + c
obs
o
The intercept, o, in the plot of k
vs. (L) was about 0.4
g
.
obs
X 10~ /sec. when the linear plot wa's'extropolated to (L) - 0.
Thus, the observed rate constant is not directly proportional
to very low leucine concentrations, as equation (19) would
predict.
In comparing these results to those obtained by others,
the following observations are pertinent.
The rate of
aldimine formation from the carbinolamine of pyridoxal and
alanine was dependent on alanine concentration at each pH and
ionic strength studied (Fleck and Alberty, 1962).
An equation
entirely analogous to (l9a) fitted the observed rate constants.
However, no proportionality between the reaction rate and
alanine concentration was found for the rate-determining
step in the transamination at pH 8 (ionic strength of 0.05 M
in sodium acetate, 25°).
The same reaction was studied at 100°, and the observed
rate constants were directly proportional to alanine concen­
tration between 0.2 and 0.4 M in a 1 to 1 acetic acid-acetate
166
buffer (Blake et ad.^ I963). The linear plots all extrapolated
through the origin (ala conc. = O).
Kinetics of the non-
enzymic transamination reaction, in general, has been found to
be first-order with respect to both pyridoxal and amino acid
concentrations (Bruice and Topping, 1963dj Blake et
Banks et
, 1963;
, 196I ).
The dependence of the rate constant on alanine concentra­
tion was not an effect of changing activity coefficient (Fleck
and Alberty, 1962). None of the parameters of the kinetic
expression was dependent on ionic strength adjusted with sodium
acetate, so that the total carboxylate concentration was con­
stant. Therefore, the dependence of the rate constant on
alanine concentration was due neither to changes in ionic
environment of the reactants nor to a specific dependence on
carboxylate ion or sodium ion concentrations. This thorough
study of the effect of ionic strength on the rate of nonenzymic
transamination by Fleck and Alberty (1962) Indicated that rate
constants of these reactions do not depend on ionic strength.
This conclusion was independently reached by Banks _et
(1961).
,
For these reasons and from the observations in pre­
liminary experiments that simple inorganic salts have no effect
on the rates, ionic strength or its adjustment to a constant
level was considered irrelevant in the model transamination
systems studied here.
167
Nonenzymic Transamination of Leucine with Pyridoxal Phosphate
Stoichiometry and kinetics
Loss in pyridoxal phosphate (PLP) was followed up to 24^
decrease in PLP at the optimum pH by the ethanolimine method
and checked by the quinolylhydrazone (QH) method (although the
keto acid data were inconsistent.) Although these reactions
followed simple second order kinetics initially (to about 25^
reaction), equilibrium conditions (correction for equilibrium
concentration of PLP) would have to be taken into account if
the reactions were followed for a longer period of time.
Since the reaction rate can be increased by increasing one
of the reactants, the best kinetic studies were made under
pseudo first-order conditions with respect to the pyridoxal
analog. The limiting reactant concentrations for this con­
dition were the solubility of leucine (0.l6 M), and the
minimum concentration of pyridoxal analog to effect a reaction
of easily measurable rate (O.Ol M for PLP).
The reactant con­
centrations used (leucine to pyridoxal phosphate) were 5^7-5
and 10 to 1. The rate of keto acid formation^ as measured by
the QH method and (checked once by the DNPH method at rela­
tively high keto acid concentration), was close to first-order
kinetics by the Guggenheim method, even for approaching equi­
librium conditions.
The rate of decrease in PLP, on the other hand, was only
close to first-order kinetics for probably up to 60% of the
168
PLP decrease.
After that, as equilibrium was approached, side
reactions Involving the pyridoxal analog may become significant
as previously suggested for deoxypyridoxal.
Although the rate
of PLP decrease was about twice as great as the rate of ketoisocaproate (KIC) formation at pH 4.1 these rates became
practically identical at pH 6.8 indicating that any side
reactions of PLP predominate at lower pH's in the acid range,
at least.
Although the total concentrations of ketolsocaproate and
pyridoxal phosphate analyzed decreased gradually in reaction
solutions, a control solution containing only these two carbonyl components, was quite stable during a five-day period
at room temperature. Very little loss of PLP or KIC occurred,
as analyzed by the qulnolylhydrazone method (Table 13). This
method also gave a linear correspondence between A^gg ^ (pyri­
doxal phosphate) and
^ (keto acid) in a reaction solu­
tion from start of the reaction (t = O) to equilibrium (t = e)
(Figure 24).
Table 13. Stability of pyridoxal phosphate (PLP) and keto­
lsocaproate (KIC) in a control solution at room
temperature
Milllmolar concentrations
PLP
KIC
At time 0
15.3
3.77
After 5 days
15.2
3.72
Figure 24.
Linearity of absorbancy at 305
(QH - KIC) and absorbancy at 400
nya (QH - PL analog) during nonenzymlc transamination of leucine
with pyridoxal analog, as measured by the modified QH method, from
Start of the reaction (t = O), buffered or unbuffered, to equi­
librium (t = e)
Initial conditions:
75 mM leu
10 mM PLP
0 to 1.0 M acetate buffer, for several experiments
pH 4.6-5.1
25°
170
171
Rate of reaction as a function of pH
Apparent second-order rate constants were determined for
unbuffered reaction solutions of leucine and pyridoxal phos­
phate (PLP) and plotted as a function of pH (Figure 25). The
optimum reaction is at about pH 4, with the rate falling off
much more sharply on the acid side than on the alkaline side.
At pH 8.5 little reaction could be detected after seven days,
although the analytical data were somewhat inconsistent.
Pseudo first-order rate constants for the transamination
of leucine with PLP were found to decrease with increasing pH
in the pH range of 3*9 to 6.8 in unbuffered solutions (Table
l4). By trial and error the equilibrium constant was calcu­
lated to be 4.1 for the system at the lower pH values, and this
indirectly-determined equilibrium position was taken into
account in calculating the corresponding rate constants.
The
Guggenheim method was used for calculating the rate at pH
6.8, but cruder methods were used to estimate the rates at
in-between pH values, as indicated in Table l4.
Intramolecular (Acid-Base) Catalysis of
IjJonenzymlc Transamination of Leucine with Pyridoxal Phosphate
Pseudo first-order rate constants for the formation of
ketoisocaproate in the nonenzymlc transamination of leucine
with pyridoxal phosphate (PLP), were compared with those for
the corresponding reaction with deoxypyridoxal (DPL) in unbuf­
fered solutions in the pH range of 4.1 to 6.8. The rate con­
stants of reactions with PLP were 3 to 4 times greater than
those with DPL. (Compare k^ values. Figure 26).
Figure 25.
Apparent second-order rate constants (k X 10 /M/sec.) for nonenzymic transamination of leucine with pyridoxal phosphate as
a function of pH
EOA analytical method for PLP decrease from initial concentrations, 0.02 M =
(PLP) = (leu), unbuffered J 25°C
(Size of rectangles around some points indicate magnitude of average error; also
value of k estimated at pH 8.5)
173
0)
-lO
-J-
••OJ
00
\0
'
J"
,_1A1 g O I x ^ ^
174
Table l4. Pseudo first-order rate constants for the nonenzymlc transamination of leucine with pyridoxal
phosphate in unbuffered solutions at 25°C
Initial mM
pH
(PL?)
(leu)
3.9
20
100
X 10^/sec.
Analytical
method
Kinetic
method
X:
KIC
-PLP
EOA
OH
log(Xg/Xg-Xj)^
4.5
9.49
±0.27
4.1
10
75
EOA
OH
It
It
4.4
20
100
EOA
It
6.30
3.40 6.98
3.94
5.47
5.8
a.0.3
7.73
+0.24
5.1
10
50
OH
log(Ae - AQ)
2.92
3.94
2.0
3.48
3.3
1.25
2.45
(^e • ^t)
EOA
QH
(ty/tx)ky
dA/dt X 23.5 mM KIC
(A-A^) 1.7
6.2
10
50
QH
6.4
10
50
QH
It
log X
6.8
10
50
QH
4.0
2.3--3.1
2.8
2.0
log A405 raplog A23Q
2.0
log (A' - A)
1.25
^Using Kg = 4.1 = (PMP)(KIC)/(PLP)(leu)
1.13
Figure 26.
Comparison of pseudo first-order rate constants
for transamination of leucine (25°) as a
function of pH by:
A.
deoxypyridoxalJ 0.01 M
O
k (unbuffered)
0
•k
°
(estimated for 0.01 M phosphate
buffer)
B. pyridoxal phosphate, 0.01 M
O
k^ (unbuffered)
•k
(calculated for 0.01 M phosphate
buffer from k values)
B
a
Based on KIC formation by QH method
^0.05 to 0.10 M leu; assumed k
concentrations
°
and k independent of
°
k ^10^
sec
T
no
-T
I
ro
0>
I—I—I
ca
I\>
1
CP
O
1—r
CP
M
J
I
I
I
I
L
177
However, since the imine formation constants, log K ,
pH
for leucine and deoxypyridoxal were found to be greater than
those for leucine and pyridoxal phosphate (Figure 15), the
relative concentrations of imine, at each pH in the range
studied, were taken into account in comparing these relative
rates of nonenzymic transamination, using equation (20). From
equation (l8) and the preceding relationships, it can be
shown that -(d(P-L)/dt) = k K (P)(L) =+ (d(K)/dt) = k
o
i o
0
obs
(P)(L), or k. K = k
, or
1 o
obs
(20)
Furthermore, in order to compare rigorously and quan­
titatively the inter- and intramolecular catalysis of phos­
phate, catalysis of nonenzymic transamination of leucine by
pyridoxal phosphate in unbuffered solutions should be compared
with that of deoxypyridoxal in the presence of an equimolar
concentration of phosphate buffer. (In such a comparison, the
concentrations of pyridoxal analog and total phosphate should
be equal, as well as that of hydrogen ion, in both systems.)
Such a comparison was made with the available data in the
following manner.
Assuming that the phosphate buffer cata­
lytic constant for transamination of deoxypyridoxal with excess
leucine is about twice that for the buffer catalytic constant
for the pyridoxal phosphate reaction (on the basis of the
comparison of the acetate buffer catalytic constants for the
two reactions at pH 4.2, Table
20), and assuming that this
ratio does not change significantly from pH 4 to 6.5, then
178
using equation (21) and points for curve A, Figure 26, the
rates for the deoxypyridoxal reaction catalyzed by 0.01 M
phosphate may be estimated for this pH range.
(21)
k ^ = k + k (s), where
Obs
o
. B
(B) =
total buffer (phosphate)
concentration, k is the buffer catalytic constant, k is the
B
o
rate constant for unbuffered solutions.
The ratio of the observed rate constants, (k , PLP/k
o
,
Ob s
DPL), for the unbuffered reaction catalyzed by 0.01 M pyridoxal phosphate to that for the reaction catalyzed by 0.01 M
phosphate buffer and 0.01 M deoxypyridoxal (estimated as
described above) was about 2^ to 4^, between pH 4.1 and 6.8
(Table 15). The gradual Increase of this ratio with pH,
except for a peak at pH 4.2, may have been a result of a
change in the ratio of buffer catalytic constants for the
two reactions with pH, i.e., the second assumption, above,
probably was not good enough for an explanation of this pH
effect with the available data.
From the ratios of calculated k^ values for PLP and DPL
as a function of pH (Table 15) it can be seen that the leucine
imine of PLP yields keto acid at rates between 5 to 11 times
faster than that of DPL in the pH range 4 to 7. This ratio
also Increased with pH, except for an optimum again at pH 4.2.
The phosphate group at the 5-position of pyridoxal indeed
catalyzed the conversion of the leucine imine to ketoisocaproate and pyridoxamine phosphate. The explanation of the
pH effect was sought in terms of the catalytic activity of the
179
Table 15. Comparison of inter- and intramolecular catalysis
of nonenzymic transamination by phosphate®-
pH
4.1
4.2
5.4
6.0
6.5
6.8
12
10.4 8.0
6.0
2.7
MC
30
OC
JO
k ^ X loT/sec., DPL l4.4 12.5
obs
42
k X lO^/sec., PLP
34
5.0
19
12
k , PLP/k
DPL
0
obs
K /Mj DPL
0'^
PLP
2.4
3.4
7.24
8.12 14.4 17.8 27.6 52.5 91.2
2.3
2.5
7.25 9.12 13.2 20.9
DPL/Kg, PLP
3.15
3.25
1.99 1.95
k^ X loT/sec., DPL
1.99
1.54 0.83 0.53 0.29 0.11 0.03
2.5
2.7 3.0
3.2
4.4
31.7
2.09 2.52 2.88
k, X lO^/sec., PLP 14.8
16.8
4.13 3.07 1.82 0.91
0.28
k., PLP/k ; DPL
10.9
5.0
9.3
7.4
5.3
6.3 8.3
Nonenzymic transamination of leucine to ketoisocaproate,
comparing deoxypyridoxal (DPL) and pyridoxal phosphate
(PLP) as amino group acceptors in pseudo first-order
conditions (excess leucine and 0.01 M PLP vs. excess
leucine and 0.01 M DPL, catalyzed by 0.01 M phosphate;
calculated rate constants from kg values for DPL and k^
for phosphate from phosphate catalysis of reaction
®
of PLP and leucine using equation (21a); assumed K and
kobs independent of concentrations
°
ionic forms of the phosphate group of PLP.
Calculation of specific phosphate catalytic constants for
the phosphate groups of pyridoxal phosphate was made using
l80
equation (21a), where k ,
ODS
Is k for the formation of keto0
Isocaproate from leucine In nonenzymic transaminations with
pyrldoxal phosphate, k^ is that for the same reaction in the
absence of the Intramolecular phosphate catalysis, using
deoxypyridoxal in unbuffered solutions as the amino group
acceptor, (B) is the pyrldoxal phosphate concentration of
0.01 M, and k ' is the specific catalytic constant for the
B
5-phosphate ester group of pyrldoxal phosphate.
(21a) k
= k + k "(B)
0
B
The data of Table l6 were used with equations (22) and
Ob s
(23), along with the literature value for the third apparent
dissociation constant for pyrldoxal phosphate, assumed to be
that of the second dissociation of the phosphoric acid ester
group (pK'^ = 6.2+p.2; Table 1, Review of Pertinent Litera­
ture), to calculate values for k and k .
HA
A
(22) pH = pK'^ + log (A)/(HA), where (A) and (HA) are concen­
trations of the dibasic and monobasic forms of the phosphate
ester group of pyrldoxal phosphate
(23) k_(B) = k (HA) + k (A), where k and k are catalytic
B
HA
A
HA
A
constants for the forms HA and A, respectively
The values calculated for k^ and k were 2.11 and.1.15
-4/ ,
HA
A
X 10 /M/sec., respectively. The monobasic form which has one
dissociable proton and a net negative charge on the phosphate
group thus had a catalytic constant of about twice that for
the dibasic form which has no dissociable protons and two
negative charges. It was concluded that the monobasic form of
l8l
Table l6. Phosphate catalytic constants (k ') of pyridoxal
phosphate in the nonenzymic
B transamination
with leucine '"
5.4
6.0
6.5
kobs
( k , for PLP)
34.0 30.0 28.0
23.6
19.0
k X loVsec?
0
(for DPL)
12.6
6.2
4.3
pH
kg' X loV^/sec.
4.1
5.0
9.0
8.4
2.14 2.10 1.96 1.74 1.47
^Reaction conditions in Table 15
^Calculated from equation (21a)
^Data of Figure 26a, O
the phosphate group in pyridoxal phosphate is a better intra­
molecular acid-base catalyst in the nonenzymic transamination
of leucine to ketoisocaproate than the dibasic form.
Further­
more, interraolecular catalysis by a second molecule of PLP and
the inductive effect of the phosphate group may be ruled out.
Another possible explanation for this effect of the neighboring
phosphate ester group is that of lowering the activation energy
by stabilization of electronic charges favorable for the trans­
ition state.
1
182
General Acid-Base Catalysis of Nonenzymlc Transaminations
of Leucine with Pyridoxal Analogs
Although the intramolecular catalysis of nonenzymic trans­
amination of leucine by phosphate (in PLP) was much more effi­
cient than the intermolecular catalysis of phosphate (in equimolar concentration with deoxypyridoxal), the intermolecular
catalysis was increased to an even greater extent by in­
creasing the concentration of phosphate in a model system.
Intermolecular catalysis compared to intramolecular catalysis
by phosphate
Catalysis by phosphate buffer seemed to shift the pH
optimum from about pH 4 up to about 5.5 (Figure 27). Buffer
catalytic constants were calculated from the limited data
available at several pH values, near the acidic optimum for
the reaction, from equation (21).
The phosphate buffer (ex­
ternal or intermolecular) catalytic constants were three to
four times greater than the
phate.
At least at pH 5.95,
constants for pyridoxal phos­
was truly a constant, and the
simple forms of equations (21) and (23) were verified because
no cross terms, such as K^k^(HA)(A) could be justified.
Calculations of catalytic constants for the specific
buffer components present in this pH range, H^PO^- and HPO^"^
were made from the limited data of Table 17, using equations
(22) and (23), where pK is pK^ ' of phosphoric acid.
The cal­
culated constants, k
and k , Indicated that most of the
HA
A
Figure-27.
General acid-base catalysis by phosphate buffer of the nonenzymic
transamination of leucine with pyridoxal phosphate as a function
of pHa
Pseudo first-order rate constants,
X 10^/sec., vs. pH
1.0 M phosphate
O
from PLP decrease by EOA method
•k^ from PLP decrease by QH method
A
from KIC formation by QH method
Q.50 M phosphate
•k^ from PLP decrease by QH method
A k^ from KIC formation by QH method
1.0 M imidazole
A k^ from KIC formation by QH method
(Lines extending from points indicate extent of experimental error)
a
'
Initial conditions:
o
0.05 M leu, 0.01 M PLP, 25 C
184
CD
00
h-
^I
Q_
D
J-
O
O
OJ
10
u
o
O
lO
185
Table 17.
Phosphate buffer catalytic constants In the nonenzymic transamination of leucine with pyrldoxal
phosphate^
KIC formation
Decrease In PLP
k
obs
k
B
k
0
k
obs
k
B
2.6
0
pH
(B)
k
0
5.95.
0.0
2.4
2.4
0
2.6
0.50
2.4
7.0
9.2
2.6 10.6
l6.0
1.00
2.4
11.6
9.2
2.6
16.2
5.4
1.00
2.8
12.5
9.7
5.0
1.00
3.0
12.0
9.0
6.5
1.00
1.9
8.5
6.6
18.8
^Pseudo first-order rate^constants (X 10^/sec. for k
and k , ^ and k ^ ; X 10 /M/sec. for k )
°
Obs
obs
^'
B
^Initial concentrations: 0.05 M leu, 0.01 M PLP, 25°
buffer catalysis (k_) is due to the acid component , H^PO.
3
c 4
In this pH range, and only a very small portion of k^ can be
p
attributed to k^ from HPO^" . This contrasts with the finding
that the catalytic constant for the monobasic form of the phos­
phate group in PLP was about twice as great as that for the
dibasic form.
The catalytic constants for the phosphate
group of PLP (1.4 to 2.1 X lO'^/M/sec.) were 20 to 23 times
greater than those for phosphate buffer (6.6 to 9-7 X 10 ^/M/
186
sec.) in this pH range.
Also, the catalytic constant for the
monobasic form of the phosphate group of PLP (2.1 X 10""^/M/
sec.) was about 23 times that of the HgPO^" species (9 X 10"^/
M/sec.) of phosphate buffer, which was the interraolecular
catalyst in this pH range.
The apparent second-order rate constant for the decrease
in PLP, from 0.02 M, equimolar, PLP and leu, was enhanced by
a factor of 6 at pH 5.9 in 0.8$ M phosphate buffer (5.5 X 10"^
M/sec. for an unbuffered solution to 33 X 10~5/M/sec. in 0.85
M phosphate).
The catalytic constant was calculated to be
32 X 10~^/M/sec., using equation (21) under these conditions.
Qualitatively the second-order rate constant was decreased at
lower buffer concentrations (0.5 and 0.1 M).
It is Instructive to ask what concentration of phosphate
buffer is required to catalyze the formation of ketoisocaproate from leucine in excess of 0.01 M DPL in order that
the rate of this reaction will be the same as that for an
unbuffered solution of PLP and leucine under the same con­
ditions.
This prediction was made using the available data
and equating k^^^ values of equations (21) and (21a): k^ +
kg(B) = k^H- kg'(0.01 M PLP), or (B) = k^:/k (O.Ol M), and
since k '/k is 23, (B) = 0.23 M, the concentration of
B
B
phosphate buffer equivalent In catalytic activity to the
intramolecular catalytic effect of the phosphate group of
0.01 M PLP, in the pH range of 4 to 6.5.
Yet to be taken
into account is the efficiency of general acid-base catalysis
187
by phosphate buffer on the basis of its frequency of collision
with intermediate aldimine under these conditions.
Intermolecular catalysis by acetate buffer
The catalytic constant for
twice the
acetate buffer was about
values for the unbuffered systems.
Pseudo first-
order rate constants for these reactions were calculated by
various kinetic methods from different analytical data and
compared in Tables l4 and 10.
These values were generally
found to decrease with increasing pH in the range of pH 3-9
and 6.8.
Above pH 6 acetate had little effect on the rate of
the unbuffered reaction.
At least at pH 3.9 and 4.6 the rate
was proportional to the acetate buffer concentration.
The
reactions near the pH optimum could be followed practically
to equilibrium in two days when the buffer concentration was
at least 0.5 M. The equilibrium constant, K^, was graphically
found by trial and error (the best straight line for a firstorder plot corrected for equilibrium concentrations) to be
4.1, and this value was used in finding equilibrium concen­
trations for kinetic methods which required them (Tables l4
and l8).
Prom the rates determined from the ketoisocaproate formed
at pH 3.9 and 4.6 (Table l8), specific catalytic constants for
the buffer components were calculated using equations (21),
(22), and (23). The constant for acetic acid, 11.2 X 10~^/M/
sec., is twice that for acetate, 5.6 X 10~^/M/sec., in this
188
Table l8.
Pseudo first-order rate constants for the nonenzymlc
transamination of leucine with pyridoxal phosphate
in unbuffered solutions at 25°C
Initial mM
pH
3.9
X 10^/sec.
Analytical
(B),M (PLP) (leu) method
0.5
10
75
OH
Kinetic
method X: fKIC
log(
X*
)
-PLP
9.5
(Xe - Xt)*
1.0
10
75
It
OH
13.8
17.1
log(A' - A)
4.6
0.5
10
50
13.5
It
OH
9.2
8.7
—— ——
—-
(AE-A,)
20
100
It
OH
EOA
10
50
OH
OH
1
50
14.05
X
0
<
10
—
12,85
d A/dt
.23.5 mM
<
1.0
—
II
tl
13.1
10.8
log (A'-A)
log (
Xe
)
12.2
log (Ae - Aq)
11.6
12.7
M MM
————
—
(^E ————
(^e " ^t)
20
100
OH
EOA
6.8
1.0
10
50
11
M# — ^ "
22.2
—
20.4
It
OH
*U8ing Kg = 4.1 = (KIC)(PMP)/(PLP)(leu)
1.2
0.94
189
pH range, which was evidence for general acid catalysis.
Cross-terms, such as
(HOAc)(OAc), in the expression
for the observed rate constant were neglected, since
was
accounted for by the sum of k^, k^^ (HOAc), and k^ (ÔAc).
For the nonenzymic transamination of alanine and pyridoxal
studied at 100° by Blake £t
(1963), the ratio of the
catalytic constant for acetic acid to that in the absence of
acetate buffer ('^^QAC/^O^ was about J.
The catalytic con­
stant, kj^oAc^ was calculated to be O.Oll/M/sec. under these
conditions. These values at 100° were quite a bit greater
than those obtained for the transamination of leucine at 25°.
The rate-determining tautomerizat'ion of imines was not
only dependent upon pH and independent of ionic strength,
but was subject to general acid catalysis by several buffers
(Banks et al., 196I; Blake et al., 1963,* Vernon, 1964).
First-order velocity constants were proportional to acetic
acid concentration which indicated general acid catalysis
(Blake _e^
, 1963). The excess alanine used is also an
.acid, and this requirement of at least 0.2 M alanine for
their analytical method prevented detailed experiments at
lower alanine concentrations and at constant pH in order to
clarify the dependence of the rate on the nature and concen­
tration of the buffer.
In the reaction between pyridoxamine and alanine Banks
^al. (1961) observed catalysis by N,N-dimethylethanolamine
and by N,N-dimethylglycine.
The equilibrium constant for
190
ketimine formation was unaffected but the rate constant for
imlne tautomerlzation was affected by both the nature and con­
centration of the buffer. Buffer catalytic constants were
calculated from an equation entirely analogous to equation
(21).
The highest rate constant for the non-metal ion-
mediated reaction between pyridoxamine and pyruvate was about
2.6 X 10-5/sec. (pH 10, 25°C, 0.1 to 0.2 M N,N-dimethylglycine buffer) (Banks^
, 1961), and this value was only
about ten times lower than that for the reserve reaction
(between pyridoxal and alanine) at 100°C, pH 4, acetate buffer
(2.85 X lO'^/sec.) (Blake _et al.j 1963), but was ten times
higher than a rate constant reported for the latter reaction
at 25°C, pH 8.0, 0.05 M sodium acetate (2.6 X 10""^/sec.)
(Fleck and Alberty, 1962).
The stoichiometry of the nonenzymic transamination of
leucine with pyridoxal phosphate, catalyzed by acetate buffer,
was carefully studied.
Millimolar (mM) concentrations of
pyridoxal phosphate (PLP) and ketoisocaproate (KIC) were
measured at various times in solutions near the pH 4 optimum
and are compared in Table 19. The total concentration of PLP
and KIC seemed to decrease during the reaction, although a
solution of just PLP and KIC was stable over a 5-day period
(Table 13). The values in Table 19 are average from several
experiments and average deviations are indicated.
Analyses
of PLP by the QH and EOA methods are compared in Table 19 and
Figure 28, and were in general agreement. The PLP measured
by the. QH method was higher than by the EOA method early in the
191
Table 19. Stoichiometry of nonenzymic transamination of 75
mM leucine with 10 mM pyridoxal phosphate (average
values from several experiments)
pH 3'9} 1.0 M acetate buffer
by EGA by QH
Total mM PLP & KIC
mM PLP mM PLP mM KIC
PLP by
EGA
t hrs.
0
8.73
10.05 0 . 0 2
+0.15 +0.06
5.44
+0.11
2.97
+0.01
5 . 7 0 3.26
+0.14 +0.36
2 . 6 5 5.63
+0.02 +0.10
1.91 6 . 5 7
+0.03
6
18.5
24
2.28
25
+0.01
2.32
44
+ 0.07
1.76
—
48
156
pH 4.1, unbuffered
0
6
24
48
67.5
0.85
+0.10
+0.02 +0.03
2.02 6.63
+0.02 +0.03
0.65 7.31
+0.13 +0.35
0.60 7.82
+0.01 +0.02
0.59
0.0
—
——— —
—
9.94
+0.04
9.04
+0.04
7.03
10.25
+0.03
8.25
+0.09
0.0
—— ——
9.14 1.40
+0.04 +0.01
6.77 2.15
+0.01
+0.10 +0.05
5.27
4.75 3 . 4 9
3.07 4 .17
—— ——
PLP by
QH
8.75
10.07
+0.15
8.70
8.96
+0.36
8.28
+0.10
8.48
+0.06
+0.36
8.60
+0.10
8.85
+0.03
8.95
+0.03
8.65
+0.07
+0.03
9.07
+0.35
+0.35
8.67
+0.10
8.42
+0.02
8.84
+0.09
+0.09
9.94
+0.04
10.44
+0.04
9.18
+0.05
10.02
——— —
7.96
8.25
10.25
+0.03
10.54
+0.04
8.92
+0.10
8.24
7.24
Figure 28.
Reaction of pyridoxal phosphate with leucine^
Millimolar concentrations vs. time, hrs.
A
KIC by QH method
Q PLP by QH method
O
PLP by EOA method
Initial conditions: 75 mM leu, 10 mM PLP, 25°
pH 3.88, 1.0 M acetate buffer
pH 4.08, unbuffered
MILIMOLT=IR CONCENTRATIONS
o -
H
i
m
X
O)
£61
194
reactions (O to kOfo decrease in PLP, and was lower than by the
EOA method in later stages of the reaction (after 40^ decrease
in PLP).
The decrease in PLP generally exceeded the formation
of KIC.
Rates based on decrease in PLP generally exceeded corre­
sponding rates for KIC formation (Table l8).
Possible side
reactions leading to loss of PLP without the corresponding
formation of KIC have been previously discussed under stoi­
chiometric and kinetic anomalies in the reaction of leucine
with deoxypyridoxal.
Rates of leucine transamination by deoxypyridoxal and
pyridoxal phosphate were compared in buffered and unbuffered
systems at two pH values, (Table 20).
Pseudo first-order
rate constants for ketoisocaproate formation determined by
the QH method indicated that the phosphate group of pyridoxal
phosphate probably enhanced the leucine transamination at pH
4.2 (in unbuffered systems) about three times faster than did
deoxypyridoxal.
However, in 1.0 M acetate buffer at the same
pH, the transamination of leucine with deoxypyridoxal was
about two times faster than that with pyridoxal phosphate.
In comparing calculated k^ and k^ values, the rate constant
for pyridoxal phosphate was about two times that for deoxy­
pyridoxal in unbuffered systems, whereas the acetate buffer
catalytic constant for deoxypyridoxal was about three times
that for pyridoxal phosphate.
At pH 6.8 pyridoxal phosphate is an even better amino
195
Table 20.
Comparison of rate constants for transamination of
leucine with deoxypyridoxal and with pyridoxal
phosphate
pH 4.2
unbuffered l.OM acetate
k,
DPL
pH 6.8
unbuffered l.OM imidazole
X 10^ kL X 10^/1^ k 7 X icf&
sec. ° sec.
sec.
1.25°
28.0^
24.6^
0.27°
ÎTT
X IoVm^
sec.
17.5
(k^, calc. = 3.4)
PLP
4.2^
12.8®
8.6^
1.2®
5.4
DPL/PLP
2.2
2.9
—
3.1
PLP/DPL 3.4
—
—
4.4
Rate constant
ratios
Pseudo first-order rate constants from Guggenheim plots
of KIC formation (by QH method)
0.01 M initial concentration of pyridoxal analog
^Second-order rate constants based on loss of pyridoxal
analog due to transamination (by EGA method)
0.02 M initial concentration of pyridoxal analog
0.03 M initial concentration of leucine
^0.05 M initial concentration of leucine
^0.10 M initial concentration of leucine
®0.075 M initial concentration of leucine
196
group acceptor than deoxypyrldoxal In unbuffered systems. The
observed rate constant for the coenzyme was almost 4^ times
greater than that for deoxypyridoxal under these conditions.
However, its buffer catalytic constant was again much less
than that for deoxypyridoxal at this pH. In comparing the
observed second-order rate constants for the loss of pyridoxal
analog due to transamination in 1.0 M imidazole buffer, the
constant for deoxypyridoxal was about three times that for
pyridoxal phosphate.
Intermolecular catalysis by imidazole buffer
A value for the rate catalyzed by 1.0 M imidazole buffer
at pH 8.6 was shown in Figure 27 and was about 2^ times
greater than the corresponding observed rate catalyzed by
phosphate buffer.
However, under similar reaction conditions,
but with deoxypyridoxal instead of pyridoxal phosphate as the
amino group acceptor, phosphate was found by Bresnahan to be
about 300 times more effective as a catalyst (1.4. X 10~^/sec.)
than imidazole (6 X lO'^/sec.) at 1.0 M concentrations, but only
1/3 as effective (1.45 vs. 4.5 X 10"^/sec.) at 0.1 M concen­
trations of buffer.
The rate seemed to be proportional to
phosphate concentration but mysteriously decreased with in­
creasing imidazole concentration (Sheryl Bresnahan, Ames,
Iowa, Iowa State University of Science and Technology, De­
partment of Biochemistry and Biophysics.
Non-enzymatic trans­
amination reactions of leucine and alanine methyl ester with
197
pyrldoxal and its analogs. Private communication, 1964).
Conditions of pH for the maximum catalytic effect of imi­
dazole were sought.
Although Bruice and Topping (1963b) quan­
titatively measured an increasing rate from pH 7 to 10 in a
related nonenzymic transamination system (pyridoxal, phenylglycine, and imidazole), neutral pH seemed to be best for the
imidazole-catalyzed reaction of leucine with pyridoxal, fol­
lowed by the ethanolimine method.
Compared to the rate at pH
8.5, where Bruice and Topping made most of their measurements,
the rate at pH 6.8 was 4 times greater for the decrease of
pyridoxal, and only f as much pyridoxal was lost in the same
time intervals in the presence of imidazole and absence of
leucine in control solutions.
However, the amounts of keto-
isocaproate formed were about the same at these pH values.
At pH 8.5 there was no increase in the rate when leucine con­
centration was increased, which would be expected if pyridoxal
is completely converted to imine with leucine at this pH.
There was also relatively more pyridoxal lost in control
solutions than there was in reaction solutions of pH less
than 6. It was concluded that the imidazole-catalyzed re­
action was too complex at pH less than 6.
Sheryl Bresnahan in this laboratory also found that the
rate of the reaction of 0.03 M leucine with 10"^ M deoxypyridoxal catalyzed by 1.0 M imidazole was greater at pH 6.6 than
at pH 8.6 (pseudo first-order rate constants, calculated by the
Guggenheim method: 6.18 fO.87 X 10~^/sec. at pH 6.62 from in­
198
creasing absorbancy at 325 mp and decreasing absorbancy at 390
my, 0.57 -0.02 X 10"5/sec. at pH 8.6, from increasing absorb­
ancy at 315 mji and decreasing absorbancies at 387 and 273 np;
1.37 X 10~^/sec. at pH 10.0, from decreasing absorbancy at 430
m^). These findings that a pH optimum for the general acidbase catalysis by imidazole is closer to neutral pH than in
the alkaline range, as Bruice and Topping suggested, seem
reasonable in the light of the pK value for imidazole being
6.95 at 25°C.
Above pH J, the nonenzymic transamination re­
action increased in rate for the unbuffered system of deoxypyridoxal and leucine, and perhaps, Bruice and Topping observed
the same effect in the alkaline range with imidazole.
The decreases in deoxypyridoxal and pyridoxal phosphate
in systems containing leucine and imidazole or phosphate
buffer, for approximately equimolar reactants, followed
second-order kinetics. The reactions were followed to 3045^ loss of pyridoxal analog, as measured by its ethanolimine,
to at least 50^ approach to equilibrium.
Pyridoxal analogs containing the 5-hydroxymethyl group
predominate in aqueous solution as internal hemiacetals in­
stead of as free aldehydes, as observed from spectra (Figure
2j Table l).
Spectra of deoxypyridoxal are evidence that
this analog exists in aqueous solution, mainly as a free or
hydrated aldehyde, which forms are much more reactive toward
amino acids than the internal hemiacetal form of pyridoxal.
Transamination between pyridoxal phosphate and glutamate pro­
199
ceeded closer to completion than that of pyrldoxal and glu­
tamate, because of the cyclic hemiacetal form of pyrldoxal
(Metzler and Snell, 1952b). It was desirable to compare the
reactivities of pyrldoxal, deoxypyrldoxal and pyrldoxal phos­
phate and thus measure the effect of hemiacetal formation in
the present nonenzymic transamination systems.
The basic character of the pyridine nitrogen and its role
in the catalysis was also considered.
Electron withdrawal
from the alpha-carbon atom of the amino acid residue in the
imine intermediate should be facilitated by a positive charge
on the ring nitrogen. The protonated ring nitrogen atom in
pyridoxylidene valine was estimated to have a pK of about
5.9 (Figure l4). However, the synthetic N-methyl analog of
pyrldoxal bears a positively-charged ring nitrogen indepen­
dent of pH. Its imines were thus expected to react faster
than those of pyrldoxal, especially at pH values above 6.
It became apparent that these relative rate comparisons
I depended on the kinetic method used. For example, if based
on calculated Initial rates of pyrldoxal analog lost in 1 M
imidazole at pH 6.8, the rate ratios were 22 for DPL, 5.5
for PLP, 3 for PLM, relative to 1 for PL. Whereas, if based
on a 3'5fo loss of pyrldoxal analog (the amount of pyrldoxal
lost in 5 hours) the rate ratios of inverse reaction times to
achieve this amount of reaction were 16 for DPL, 3 to 4 for
PLP, 1.6 for PLM, relative to 1 for PL (Figure 29). The equi­
librium position seemed to be at about 50^ loss of pyrldoxal
Figure 29. Relative transamination rates of four amino
group acceptors with leucine, catalyzed by 1.0
M imidazole buffer^.
t hrs. for 3.5#
transamination of
pyridoxal analog*^'°
Relative rates
5A =
= V'^Pl
DPL
0.30
PLP
1.2-1.7
PLM
3.0
1.6
PL
5.0
1.0
^Initial conditions:
l6
3-4
0.03 M leu, 0.02 M pyridoxal
analog, pH 6.8, 25°
^At 3.50 decrease in pyridoxal analog, the amount of
pyridoxal lost in 5 hours in the presence of leu and
imidazole, minus the amount lost in the same time in
the absence of leu and presence of imidazole
^Measured decreases in pyridoxal analog by ethanolimine method and, after correcting for loss in PL and
PLM due to imidazole interaction, calculated per cent
pyridoxal analog lost due to transamination with
leucine
201
Deoxy pyndoxal(DPL)
Pyridoxcil
10-
me+hochloride (PLM)
TIME , HRS.
202
analog, under these second-order conditions.
Second-order
rate constants were calculated for the reactions with DPL
(17.5 ±1.1 X lO'V^/sec.) and with PL? (5.36 +0.22 X lO'V^/
sec.), and the rate ratio, k, DPL/k, PLP - 3.3 was obtained
under these conditions.
An attempt to compare pyridoxal and its N-methyl analog
in an imidazole system at pH 4 indicated a much slower re­
action, which was difficult to measure (only 3 to 4^ pyri­
doxal was lost in 84 hours). The best reactions catalyzed
by imidazole seemed to be above pH 6, unfortunately.
These results may be explained on the basis that buffer
catalysis vias a greater effect than the intramolecular cata­
lysis by functional groups substituted at the 5-position of
pyridoxal.
These intramolecular effects may only be observed
in the absence of buffer or in very low buffer concentration.
These catalytic groups, such as the phosphate ester in pyri­
doxal phosphate, interfere with buffer catalysis, resulting
in lower buffer catalytic constants than those for deoxypyridoxal, which has no functional group In the 5-position.
Comparisons of transamination rates of pyridoxal phos­
phate, and other pyridoxal analogs, with deoxypyrldoxal should
take into account the relative amounts of aldlmine inter­
mediates present (from experimentally determined imlne for­
mation constants), unless these intermediates are present at
low steady state concentrations (but this Is not the case for
these reaction conditions where relatively high reactant con-
203
centratlons are used). The rates of conversion of the aldlmlnes of different pyrldoxal analogs to products would yield
a fair comparison of the relative, transamination rates ef­
fected by Intramolecular catalysis.
Significant decreases of pyrldoxal and of pyrldoxal methochlorlde In control solutions containing 1.0 M Imidazole were
measured by the ethanollmlne method. These decreases were not
reproducible nor could they be predicted.
Corrections for
these changes caused by Imidazole were made to obtain net pyrl­
doxal decreases which were due to transamination with leucine.
However, no such effect of Imidazole of deoxypyrldoxal and
pyrldoxal phosphate were observed. The stability of these
pyrldoxal analogs In the presence of Imidazole suggests that
the hydroxymethyl group In position 5 of pyrldoxal and Its
N-methyl analog Is necessary for this effect.
Perhaps, Imidazole somehow affects the equilibrium of
these latter compounds to favor their hemiacetal forms and to
inhibit the reaction with ethanolamine, which presumably
reacts with the free aldehyde forms in yielding ebhanolimines
quantitatively.
The gradual decreases of these two pyrldoxal
analogs in these control solutions could be accounted for on
the basis of slow interactions with imidazole, such as a shift
in equilibrium or a formation of a condensation product re­
quiring a hydroxymethyl group. As far as this author can de­
termine, Bruice and Topping did not report the use of such
control solutions of pyrldoxal and Imidazole in their studies
204
of the kinetics of the imidazole - catalyzed reaction between
pyridoxal and phenylglycine.
Bruice and Topping, who have reported kinetic studies of
the imidazole-imidazolium ion concerted general acid, general
base catalysis of the transamination of phenylglycine with
pyridoxal in a recent series of publications, have claimed
that the uniqueness of imidazole as a catalyst is in its
solubilization of amino acids by complexing with them, as well
as with the imine intermediates (1963d). Phenylglycine, an
unnatural amino acid, is less soluble in water than leucine.
But when its solubility increased in solutions of imidazole
buffer, Bruice and Topping suggested the formation of a
complex between imidazole and the phenyl group of the amino
acid (1963c).
However, DePrenger, in this laboratory, also
found that leucine, an aliphatic amino acid, had greater
solubility in solutions of imidazole buffer than in water
alone.
No significant difference was found between observed
rates of nonenzymic transamination of leucine and of phenylglycine (Donald DePrenger, Ames, Iowa, Iowa State,University
of Science and Technology, Department of Biochemistry and
Biophysics.
A study of rates of transamination of pyridoxal
analogues catalyzed by Imidazole.
Private communication.
1964). However, the fact that a higher maximum wavelength
was found for the solution of leucine, pyridoxal, imidazole
(410 to 415 mp) than the corresponding solution of phenyl-
205
glycine (395 mp) under the same conditions indicated a greater
extent of imine formation in the leucine solution.
Since the
rates were the same, the imine of phenylglycine and pyridoxal
must be more reactive than that of leucine and pyridoxal.
This difference may be explained by the difference in electromeric effects of a phenyl group and an alkyl group on the
bond-breaking process at the alpha-carbon atoms of the amino
acid residues in these imines.
However, at high amino acid concentration DePrenger found
that the reaction with leucine proceeded much faster, but fol­
lowed zero-order kinetics compared to the slower reaction with
phenylglycine which still followed pseudo first-order kin­
etics.
But when the pyridoxal concentration was tripled so
that the molar ratio of leucine to pyridoxal was 67 to 1,
instead of 200 to 1, a pseudo first-order rate constant was
calculated as I.5 X 10 ^/sec., significantly smaller than that
calculated for the reaction in which only a 2 to 1 molar ratio
of leucine to pyridoxal existed at 10"^ M concentrations (3.7
X 10~^/sec.). It is not understood why the observed rate de­
creased with an Increase in leucine concentration.
The reaction of leucine with deoxypyridoxal under these
conditions with leucine in excess by a factor of 200 times was
found by DePrenger to be further complicated by a slow for­
mation of imine, which was followed by the absorbancy at 4lO
mp increasing to a maximum in 12 minutes after mixing the
reactants.
After this time the rapid decrease in absorbancy
206
followed zero-order kinetics.
Louise Hodgin and Sheryl
Bresnahan in this laboratory also independently obtained zeroorder kinetics for the same reaction. (Increasing absorbancy
at 317 mp and decreasing absorbancy at 415 or 387
were
directly proportional to time.)
Bruice and Topping (1963a) found that increasing the
imidazole concentration from 0.1 to 1.8 M increased the rate
of ketimine formation from lO"^ M reactants at pH 8.6 by a
factor of 200 times (k
increased from 8 X 10~^/sec. to 1.5
obs
X 10" /sec.). Furthermore, the rate was found to depend upon
the square of the imidazole concentration up to a maximum of
about 1.8 M, supporting their claim that two molecules of
imidazole complexed with aldimine in the acid-base catalysis
of the rate-limiting step.
When the concentration of phenyl-
glycine was doubled for the same reaction conditions in 1.8
M imidazole the observed rate constant was further increased
-4 /
to 2.0 X 10 /sec.
Under these same conditions DePrenger in this laboratory
-4
observed a rate constant of only 0.37 X 10 /sec.j 5i times
less than that reported by Bruice and Topping.
Perhaps, in
the calculation of their rate constant, Bruice and Topping
only considered the formation of ketimine in "correcting"'
their observed rate constant, since the relative slowness of
the conversion of ketimine to products prevented a precise
determination of the final equilibrium owing to complicating
side reactions in this final step (1963a; also see rate equa-
207
tlon. Review of Pertinent Literature). But DePrenger simplycalculated an observed rate constant for the over-all trans­
amination reaction.
DePrenger also found no significant difference between
rates measured at 25° and at 30°C (the latter being the tem­
perature used by Bruice and Topping).
Nor was there any
effect of Ionic strength from 0.05 M KCl, which Bruice and
Topping and others so carefully controlled in their model
systems.
Finally, Donald DePrenger, Louise Hodgin and Sheryl
Bresnahan in this laboratory compared rate constants for the
nonenzymlc transamination of these amino acids with several
pyridoxal analogs catalyzed by imidazole buffer and followed
by spectral methods.
In 1.8 M imidazole, pH 8.6, pyridoxal
(3.67 X 10~^/sec.) reacted slightly faster than pyridoxal
phosphate (3.4 +_0.2 X lo'^/sec.) with phenylglycine, in which
there was a 2 to 1 molar excess of amino acid in both cases.
-5 /
The latter constant may be compared with 3.5 X 10 /sec.,
(as measured by decreasing absorbancy at 278 m;;) in which
there was a 67 to 1 molar excess of leucine.
Nor did the
leucine imine of deoxypyridoxal (5.8 X 10~^/sec., as measured
by decreasing absorbancy at 426 mp) react much faster than
that of pyridoxal (5.5 X 10~^/sec., as measured by decreasing
absorbancy at 4lO m^i).
For this latter comparison there was
a 500 to 1 molar excess of amino acid in these reactions.
Again, it is not understood why the rate constants have de-
208
creased with an increase in amino acid concentration.
Appar­
ently at pH 8.6 in high imidazole buffer concentration the
hemiacetal form of pyridoxal did not decrease the reaction
rateJ as compared with the rates for deoxypyridoxal and pyri­
doxal phosphate, although a large effect of the hemiacetal
form was observed at pH 6.8 (Figure 29).
Decreasing the concentration of imidazole to 1.0 M an
average rate constant for the reaction of PLP with a 200 molar
excess of leucine was calculated to be 7.3 jfO.S % 10 ^/sec.
(as measured by decreasing absorbancies at 4l4 and 275 myi
and increasing absorbancy at 322 mp).
This was compared with
a value of 2.5 X 10~^/sec. (as measured for KIC formation by
the QH method), in which there was only a 5 to 1 molar ratio
of leucine to PLP.
These values obtained in this laboratory
may be compared with 3 X lO'^/sec. obtained by Bruice and
Topping (1963d) for the formation of ketimine from equimolar
pyridoxal and phenylglycine under the same conditions of pH
and imidazole concentration.
A striking effect of a positive charge on the pyridine
nitrogen atom was observed by DePrenger In the laboratory.
In 1.8 M imidazole, pH 8.6, the leucine imine of pyridoxal
methochloride (58 X 10~^/sec., as measured by decreasing
absorbancy at 385 m)a) reacted ten times faster than the imine
of pyridoxal (5.5 X 10 ^/sec., as measured by decreasing
absorbancy at 410 irya). (The wavelengths of maximum absorbancy
of these imlnes differ, as expected from the differences in
209
Ionic forms.) This lends further support to the mechanism
which is favored by a positive charge on the pyridine ring
nitrogen atom of the imine.
At pH 8.6 the imine of pyridoxal
and leucine has dissociated the proton from the pyridine
nitrogen atom, whereas that of pyridoxal methochloride, of
course, retains a positively-charged pyridine nitrogen atom.
However, a quantitative comparison of the reactivities
of these pyridoxal analogs should take into account their
relative aldimine and hemiacetal formation constants under
these conditions.
In the latter comparisons DePrenger found
that the apparent molar absorbancy index of the pyridoxal
methochloride imine was only about -ç that of the imines of
pyridoxal or of deoxypyridoxal.
This markedly decreased
stability of the PLM imine,despite the estimated I.7 times
greater concentration of PLM free aldehyde than that of PL,
emphasizes that the positive charge on the pyridine nitrogen
atom is important in producing a high degree of reactivity.
It is quite likely that the PLP coenzyme may be attached to
the protein in such enzymes through hydrogen bonding at the
pyridine nitrogen atom.
From these results it is suggested
that such attachment would have a profound effect on the
reactivity of the imines which exist as intermediates in the
catalytic process.
I
210
Metal Ion Catalysis of Nonenzymic Transamination of
Amino Acids with Pyrldoxal Analogs
Metal ion enhancement of general acid-base catalysis
The general acid-base catalysis of nonenzymic trans­
amination of leucine with pyridoxal analogs was further en­
hanced by the addition of metal ions.
Near the pH 4 optimum
the additive effects of acetate buffer and cupric ion were
measured by the following calculations.
The rate constant, k^', for the reaction of 0.1 M leu­
cine (leu) with 0.01 M deoxypyridoxal (DPL) in unbuffered
solution at pH 4.2, was calculated to be 3.4 X 10 ^/M/sec.
from the
' value of 2.50 X lO'^/M/sec. obtained for 0.05 M
leu multiplied by the factor 72/53, the ratio of k^' values
for the two leucine concentrations at pH 7.2 (Table 12). The
pseudo first-order rate constant for the reaction with 0.1 M
leu in unbuffered solution at pH 4.2 is predicted to be 3.4
X lO'^/sec. by equation (l9a).
The observed pseudo first-order rate constant for these
reaction conditions in the presence of 1.0 M acetate buffer
was 2.8 X 10~^/sec., which was an eight-fold increase over
that of the reaction in unbuffered solution. (These values
were for KIC formation measured by the QH method.) The cata­
lytic constant for acetate buffer, obtained by use of equation
(20), was 2.46 X lO'^/M/sec.
The addition of 1.0 mM Cu (ll) to this model system almost
211
doubled the observed rate constant to 4.7 X 10 ^/sec. This
constant, however, was calculated Indirectly from a secondorder rate constant of 1.2 X 10~^/A
/sec. multiplied by
3O5 MP
the maximum absorbancy at 385 m;ji of 0.393.
This wavelength
is the absorption band for the cupric chelate of the aldimine
of leucine and DPL, and loss of absorbancy at this wavelength
followed simple second-order kinetics (straight-line slope
for l/A vs. t) after it was found that a Guggenheim plot con­
caved sharply downward.
The maximum absorbancy at 385 mp.
was reached at about ten minutes after mixing, indicating
slow chelate formation with the aldimine. From equation (24),
the catalytic constant for cupric ion, k
,
, was calcuo
Cu (II)
lated as 19 X lO'vM/sec. (24) k
= k ' (leu) + k (B) +
Ob s
o
3
"•cu (ID
The catalytic constant for cupric ion was about 700 times
greater than that for acetate buffer under these conditions.
The observed rate constant of 4.7 X 10 ^/sec. may be compared
with that of 3 X 10~^/sec. for the reaction between pyrldoxal
and phenylglycine catalyzed by 1.0 M imidazole at pH 8.6
(Bruice and Topping, 1963d), which was not affected by alumi­
num ion, and with that of 2.6 X 10~^/sec. for the reaction of
pyrldoxal and alanine in N,N-dimethylglycine buffer at pH 10,
which was further catalyzed by cupric ion by a factor of 100
(Banks et al., 1961).
Catalysis of the reaction of pyrldoxal with leucine was
enhanced by about three times when 0.02 M zinc salt was in-
212
eluded with the system buffered with 1.0 M Imidazole at pH
6.8.
This contrasted with the conclusion reached by Bruice
and Topping (1963a) that metal ions do not affect imidazolecatalyzed reactions, from their observations that aluminum
ion did not affect the observed rate at pH 8.6 catalyzed by
1.8 M imidazole.
Comparisons were made also in the zinc-imidazole system
between the reactivities of pyridoxal (PL), pyridoxal methochloride (PLM), deoxypyridoxal (DPL), and pyridoxal phosphate
(PLP) as amino group acceptors.
Relative rates were based on
calculations of initial rates of decrease in pyridoxal analog,
as measured by the ethanolimine method.
All of these re­
actions were 0.02 M in pyridoxal analog, 0.03 M in leucine,
pH 6.6 to 6.8.
In addition, 1.0 M imidazole, 0.02 M zinc
ion with ionic strength adjusted to 0.5 M with sodium acetate,
or 1.0 M imidazole plus 0.02 M zinc ion were included in these
model systems.
The relative rates in 1.0 M imidazole buffer
were compared in the previous section.
In the systems containing zinc ion under these conditions
a white precipitate formed in the reactions with DPL at time
zero, and no further attempts to compare the zinc catalysis"
of the reactions with DPL and PLP were made.
It was quite
likely that in solutions of relatively high concentrations of
imine (DPL or PLP with leucine) the resulting zinc chelate of
these imines formed to an extent exceeding their solubility.
But in solutions of relatively low concentrations of imine
213
(PL or PLM with leucine) rate comparisons were made.
Using equation (9), initial rates were converted to
second-order rate constants:
-8
dP/dt was 3 X 10" /M/sec. or
kg = 5 X 10 ^/M/sec.j for PL and leu, in the zinc ion-sodium
acetate system.
These values were 1.7 times greater than
those for PLM and leu in a corresponding system.
But in the
g
zinc ion-imidazole buffer system, dP/dt was 12 X 10
M/sec.
or kg = 20 X 10"5/M/sec. for PL and leu, and the corresponding
values for PLM and leu were 1.6 times greater.
Reasons for
these apparently conflicting results are not at all obvious.
Blake e_t ^., (1963) also observed enhancement of acidbase catalysis by aluminum ion.
An acetate buffer (0.1 M
acetic acid, 0.1 M acetate) doubled the observed rate constant
of the reaction between alanine and pyridoxal at 100°C compared
to the reaction in unbuffered solution of the same pH with
other conditions held constant.
When 5 X 10"^ M aluminum ion
was included in this system with the acetate buffer, the rate
constant was tripled.
Qualitative observations of these
effects were made earlier by Metzler and Snell (1952b).
Metal ion catalysis of nonenzymic transaminations of amino
acids with pyridoxal and with pyridoxal methochlorlde
Attempts were made to measure the effect of N-methylation in pyridoxal on the nonenzymic transamination of amino
acids catalyzed by metal ions.
Rate measurements were based
on millimolar concentrations of pyridine aldehyde lost or of
keto acid formed after 10 and 30 minutes near the boiling
214
point of water at two pH values measured at 25° and controlled
by 0.2 M buffers.
The transamination of leucine with pyri-
doxal methochloride (PLM) indicated up to l|- times more
changes in concentrations of reactants and products than that
with pyridoxal (PL) at pH 4.8 (Table 21).
However, concentration changes in the reactions of
serine with PL and PLM were about the same.
Rates of for­
mation of products analyzed as keto acid from serine were
much greater than the corresponding rates of decrease in
pyridine aldehyde. This latter anomaly may be due to the
side reactions of serine with pyridoxal, such as /3-elimination and aldol cleavage, which lead to different carbonyl
products, other than pyruvate via transamination.
And the
dinitrophenylhydrazone product from serine was assumed to
be pyruvate, since the keto acid formation was calculated
from a pyruvate standard curve.
Analytical methods, and pro­
cedures of Metzler and Snell (1952b) were used in these
experiments.
Under the conditions of a pH near neutrality the re­
actions are much slower than at pH 4, in agreement with
earlier observations that the pH optimum for aluminum ioncatalyzed transamination is about 4.5 (Metzler and Snell,
1952b).
However, as expected, the leucine transamination
rates were even higher with the N-methyl analog than with
pyridoxal at the pH above the pK of the pyridinium group in
the pyridoxal imine. In this case, the reaction with the
215
Table 21.
Comparison of reactions of pyridoxal (PL) and
pyrldoxal-N-methochloride (PLM) with amino acid
at 100°C catalyzed by aluminum ion (alum.) at pH
4.8 (20OC): 0.2 M in NH.OAc + HOAc, 0.3 M ionic
strength
mM changes in 10'
in 30'
Pyridine
aldehyde
lost
Keto acid
formed
Pyridine
aldehyde
lost
PL + ser
2.3
8.3
2.5
11.8
+ leu
3.9
3.7
5.7
5.4
PLM +ser
2.1
8.9
2.5
10.7
+ leu
6.0
4.7
6.8
5.6
1 mM alum.
+10 mM reactant
Table 22.
Keto acid
formed
Comparison of reactions of pyridoxal (PL) and
pyridoxal-N-methochloride (PLM) with amino acid
at 100°C catalyzed by aluminum ion (alum.) at pH
6.6 (20°C): 0.2 M in phosphate buffer, 0.3 M
ionic strength
mM changes in 10'
Pyridine
aldehyde
lost
Keto acid
formed
Pyridine
aldehyde
lost
Keto acid
formed
PL + ser
0.0
0.0-0.1
0.1
0.0
+ leu
OC
o
1 mM alum.
+ 10 mM reactants
in 30'
0.3-0.6
1.8
1.0
PLM + ser
0.0
0.0-0.05
0.0-0.3
0.13
1.26
0.82
2.7
2.1
+ leu
216
positively charged imine of the N-raethyl analog proceeded to
between l|- to 2|- times farther in 10 and 30 minutes than with
the imine of pyridoxal (Table 22).
Also, these rates were
clearly greater than those of the serine reactions.
These
results are also in harmony with the proposal that the Nprotonated (or M-methylated) form of the metal chelate is the
reactive form (Johnston et al., 1963).
These rates of leucine transamination were expressed as
initial rates, dP/dt, for the concentration changes after 10
minutes reaction (Table 23).
As usual, the rates of decrease
in pyridine aldehyde were greater than the rates of keto acid
formation.
Table 23.
Comparisons of reactions of pyridoxal (PL) and
pyridoxal-N-methochloride (PLM) with amino acid
at 100°C catalyzed by aluminum ion (alum.):
calculated initial rates, dP/dt X lO^/M/sec., of
leucine transamination from Tables 21 and 22
pH
4.8
pH 6.6
(kg/M/sec.)
PL & leu
PL lost
6.5
(0.18)
1.3
KIC formed
6.2
(0.17)
0.5-1.0
(0.20)
2.1
(0.16)
1.4
PLM & leu
PLM lost
KIC formed
10
7.8
217
The extents of these reactions at pH 5 were compared with
earlier values (Table 24). It was assumed that the equilibrium
values for ^ pyridine aldehyde lost in the transamination of
leucine were 60^ for PL and "JOfo for PLMj in order to estimate
the apparent second-order rate constants from the initial
rates of Table 23.
Prom equation (9), relating initial rates,
dP/dt, to second-order rate constants, kg, these constants
were calculated using 6 mM for the "effective" or reacting
concentrations of PL and leucine and 7 mM for the reacting
concentrations of PLM and leucine.
These values were near
0.2/M/sec. (Table 23), and may be compared with the highest
rate constant obtained by Blake ^
, (1963) of 0.0025/M/
sec. for the decrease in pyridoxal, starting with 10 mM PL,
0.20 M a alanine, 0.10 M acetate, 0.10 M acetic acid, catalyzed
by 5 X 10"^ M alum, at 100°C. This comparison indicates that
a 20 times higher concentration of aluminum ion catalyzed the
reaction by 100 times under these conditions, if leucine and
alanine have about the same reactivities as amino group
donors.
The stability of pyridoxal methochloride (PLM) was
checked at 100°C, because some of its reactions with amino
acids were measured under these conditions.
The solutions
were 10 mM in PLM, 1 mM in alum., and were buffered with
either 0.2 M ammonium acetate at pH 4.7 or with 0.2 M phos­
phate at pH 6.6, 25°C.
A portion of each solution was heated
in a boiling water bath for 30 minutes.
Aliquots were then
218
Table 24.
Comparison of reactions of pyridoxal (PL) and
pyridoxal-N-methochloride (PLM) with amino acid
at 100°C catalyzed by aluminum ion (alum.) at
pH 5 (20OC): 0.2 M in NH.OAc + HOAc
1 mM alum.
+ 10 mM reactants
% pyridine aldehyde lost after 30'
PL + ser
25
PLM + ser
25
PL + leu
57
PLM + leu
68
PL + leu
. 51^
After 120' (equilibrium)
59^
PL + leu
a
Source:
Metzler and Snell, 1952b
acidified to pH 0.2 in diluting them to a convenient concen­
tration for spectral analysis.
Absorbancies of these solu­
tions were measured at 288 and 293 mp against appropriate
blanks.
At either pH, absorbancies of all solutions were
greater at 293 than at 288 mp.^ the absorption maxima of the
hemiacetal cation forms of PLM and, PL, respectively. The
ratio of absorbancies at these two wavelengths for each solu­
tion at pH 4.7 remained constant within one percent, (attri­
butable to experimental error).
This constant absorbancy
ratio was expected if the heat treatment did not lead to hydro­
219
lysis or cleavage of the N-methyl group to yield pyridoxal,
which has a higher absorbancy than PLM at the lower wavelength.
The ratio of absorbancies for each solution at pH 6.6
indicated either a maximum of three percent
loss of PLM due
to an opening of the N-methylpyridine ring or to experimental
error caused by discrepancies in the absorbancies of the
blank solutions.
But at least no significant extent of de-
methylation to pyridoxal occurred under the reaction con­
ditions (Johnston et al., 1963).
Under similar reaction
conditions the individual stabilities of pyridoxal, pyrldoxamine, amino and keto acids were checked previously
(Metzler and Snell, 1952b), and remarkable stability was
noted for most of these compounds during the relatively
short reaction times required (l to 3 hrs.) at 100°C.
Since this model transamination reaction also increases
with an increase in temperature, the rates at 25®C were only
about 2% as great as the corresponding rates obtained near
100°C (boiling water).
Metal-ion catalyzed reactions were
followed at 25°C, using the modified analytical methods
adapted from the procedures of Metzler and Snell (1952b).
In a comparison between pyridoxal and its N-methyl analog at
pH 4 the rates of A1 (Ill)-catalyzed reactions were about the
same, although the pyridoxal reaction appeared to be slightly
faster under these conditions (Figure 30). Pseudo firstorder rate constants, calculated by the Guggenheim method.
Indicated that the rates of decrease in pyridoxal (2.7 X lO"^/
Figure 30.
Comparison of nonenzymic transaminations of leucine with pyridoxal
(PL; and with pyridoxal-N-methochloride (PLM)^
mlllimolar (mM) concentrations vs. time
o
decrease in PL
o
decrease in PLM
formation of KIC from leu and PL
d
formation of KIC from leu and PLM
a
Initial conditions: 120 mM leu, 10 mM PL or PLM, 2 mM alum., pH 4.2, 0.1 M
acetate buffer, 0.5 M ionic strength, 25°C
ô"
mM
-O
4-
20
40
60
60
100
TIME(hours)
120
140
160
222
sec.) and of formation of ketoisocaproate (2.85 X 10 ^/sec.)
were about the same', although initial rates (dP/dt) indicated
a significantly greater rate of decrease in pyridoxal.
The
initial rates were found to be constant over the first five
hours at least.
These pseudo first-order rate constants were
only slightly higher than the best rate constant reported by
Banks £t
(1961) for the reaction of 0.2 M alanine with
0.01 M pyridoxal at pH 10 in 0.1 M N,N-dimethylglycine buffer
(2.6 X 1 0 "Vsec.).
Figure 10 shows that the rates of reaction between pyri­
doxal and leucine catalyzed by aluminum ion in 0.1 M acetate
buffer were about the same as those for deoxypyridoxal and
leucine in 1.0 M acetate buffer without added metal ion.
This indicates that catalysis by 2 mM aluminum ion is at
least equivalent to that by 1.0 M acetate buffer under these
conditions at pH 4.2, since deoxypyridoxal is a much better
amino group acceptor than pyridoxal.
In a similar reaction in which there was about 2-| times
less initial concentration of leucine the initial rates were
about half as great.
Equilibrium was approached after 10
days, indicated by a 52^ decrease in pyridoxal.
A control
solution containing sodium alpha-ketoisocaproate, instead of
leucine under reaction conditions, indicated no significant
changes in pyridoxal or keto acid concentrations, within
experimental error, during the first five hours at least.
Studies with zinc ion catalysis under these conditions
223
Indicated a very slow reaction even after 25 hours.
A rough
comparison between the aluminum and zinc ion catalyses at pH
4 indicated a l4^ decrease in pyridoxal in the reaction with
leucine and aluminum ion in five hours and 1% decrease in the
zinc ion-catalyzed reaction. Furthermore, 8 times more keto
acid was produced from the former reaction than from the
latter in this time.
However, zinc ion catalysis is probably one of the most
favorable metal ion systems for quantitative rate studies,
because, in comparison to other metal ions, zinc ion exchanges
rapidly with the hydrogen-bonded proton in the aldimine inter­
mediate and it has a relatively simple solution chemistry.
Furthermore, oxidation side reactions caused by zinc ion are
of little significance, and formation constants of zinc ion
with ligands are Icnown.
The only disadvantage is that zinc
is a mediocre metal ion catalyst, far below the catalytic
effectiveness of copper, iron, and aluminum ions (Metzler
and Snell, 1952b ; Gregerman and Christensen, 195^; Longenecker
and Snell, 1957). Although manganese (ll) was recently re­
ported to be one of the best metal ions for promoting the
transamination of glutamate with 3-hydroxypyridine-4-aldehyde
(Bruice, 1964), it had only mediocre activity in catalyzing
the transamination of glutamate with pyridoxal at 100° (Metzler
and Snell, 1952b; Longenecker and Snell, 1957).
Zinc ion catalysis was found to be more effective at a
higher pH, just below neutrality (Table 25; reactions 1,2,5),
224
in agreement with the findings of Longeneclcer and Snell (1957).
The reactions proceeded best when the ionic strength was ad­
justed with sodium acetate than with perchlorate or chloride
salts - another indication of general acid-base catalysis.
The gradual formation of a white precipitate as the pH was
increased and was prevented during the first day of reaction
by adding the zinc salt after mixing the other components
(forming the imine first).
Concentrations of leucine and
zinc perchlorate were varied (Table 25; reactions 2-7) to
find reaction rates comparable to those of the aluminum ioncatalyzed reactions at pH 4 (Table 25; reactions 2,5,6,8),
although reproducibility of the analytical methods used in
following these reactions was undesirable.
Under the conditions indicated near neutral pH, initial
rates of pyridoxal lost and ketoisocaproate formed were re­
producible and fairly constant, within experimental error,
during the first 8 hours of reaction (Table 26), after ana­
lytical procedures were modified (See Experimental). For­
mation of some white precipitate, a big problem in metal ion
systems especially at pH values above 5j was noticeable after
one day and prevented further reliable rate measurements
beyond the initial rates, thus making reaction order and
rate constants difficult to determine.
In 5 hours this re­
action proceeded to about 2^ and in 8 hours to about 4g^.
Or, about 1^ of the reaction, with reference to the equili­
brium point, occurred per hour in the initial reaction con-
225
Table 25. Comparisons of metal ion-catalyzed nonenzymic
transamination rates at 25°C
Initial conditions
0.3 p. salt
dP/dt after 2
hrs. (-moleSnPL/
l/sec. X 10°)
mM PL
mM Leu
(NaOAc,CIO,-,
CI-)
'+
pH
1)
10
80
10 mM Zn (II)
4.7
1
2)
20
21
20
45
45
45
20 mM Zn (II)
20 mM Zn (II)
20 mM Zn (II)
6.2
6.5
6.5
4
7
12
3)
20
20
20 mM Zn (II)
5.7
3
4)
20
30
10 mM Zn (II)
7.3
10
5)
20
20
30
30
20 mM Zn (II)
20 mM Zn (II)
6.8
6.9
6
12
6)
20
20
30
30
30 mM Zn (II)
30 mM Zn (II)
6.6
6.7
9
10
7)
20
30
40 mM Zn (II)
6.5
4
8)
10
10
120
2 mM A1 (III) 4.2
2 mM A1 (III) 4.3
15
8
22
45
9)
50
20 mM Cu (II)
5.8-6.5
65
ditions.
An approximation of the over-all second-order rate con­
stants, comparing pyridoxal and its N-methyl analog in this
zinc system, was made from the Initial rates using equation
(9). The lower reactivity of pyridoxal methochloride (3 X
226
Table 26. Zinc ion-catalyzed transamination rates®"
Hours:
P:^
1)
2)
5
8
7
PL
lost
PL
lost
KIC
formed
4.1
3.1
3.2
+0.3
+0.2
+0.6
PL
lost
KIC
formed
2.9
2.8
3.1
3.8
+0.3
+^0.6
+0.6
+2.2
3.6
+0.2
KIC
formed
^Initial concentrations: 20 raM PL'HCl
30 mM Leu
20 mM Zn (CLO^)
0.5 M with NaOAc, pH 6.8, 25°C
b
Q
Overall rates at times indicated: dP/dt X 10 M/sec.
10"^/M/sec.) may be due to the lower pH recorded for it (6.2)
as compared to the pyridoxal system (5 X 10 ^/M/sec., pH 6.8).
However, previous to this, unexplained difficulties were ex­
perienced in trying to measure a reaction between pyridoxal
methochloride and leucine in this zinc system.
A few experiments with cupric ion catalysis, which was
found to be the best metal ion catalyst of these model system
reactions (Metzler and Snell, 1952b), indicated that cupric
chelates of imines were not completely soluble at pH 6. At
this pH, the cupric ion-catalyzed reaction was between 10 and
227
15 times faster than the zinc ion-catalyzed reaction, without
considering that most of the cupric ion probably had pre­
cipitated out of solution (Table 25; Reaction 9).
A more thorough study of the cupric ion-catalyzed re­
action of pyridoxal with at least a 5-fold excess of leucine
was made at pH 5, 25°C by Dr. Keith Schmude in this laboratory.
Observed first-order rate constants were obtained from plots
of absorbancy of portions of the reaction solution in ethanolamine at 3^5 mp vs. time,.
The appearance of two distinct
regions in a plot when Cu (ll) and pyridoxal concentrations
are approximately equal may be due to slow formation of the
imine chelate.
The rate constant was not a linear function
of the concentration of Cu (ll).
The precipitate which formed in cupric ion model systems
was probably cupric dileucinate with a solubility product of
-8
about 1.5 X 10
at pH 5.
It was predicted that the maximum
Cu (ll) concentration permitted would decrease as the square
of the leucine•concentration to explain why the rate did not
increase with increasing concentrations of leucine and pyri­
doxal.
Also, oxidation of pyridoxal was probably catalyzed
by Cu (II), since about half of the pyridoxal lost in the
presence of leucine was lost in its absence.
Furthermore, the rate seemed to depend on the ratio of
concentrations of Cu (ll) and pyridoxal and was maximum when
this ratio was unity, which indicated that a simple one-toone imine chelate was the reactive intermediate.
This con-
228
elusion was also reached from kinetic studies of the reverse
reaction, that of pyridoxamine with excess ketoisocaproate,
catalyzed by cupric ion.
Zinc ion was found to be 100 to 200
times less effective as a catalyst than cupric ion under these
conditions. However, no deviation from linearity in the first
order plots was observed, indicating a rapid formation of zinc
chelate.
Measurements of absorbancies at 380 m)i as a function
of- time verified the relative rates of imine chelate formation
from these two metal ions.
Also, apparently much less zinc
chelate was formed as compared to the cupric chelate, from
these spectral studies.
The reaction rate decreased at lower
pH, as expected from the decreased stability of the inter­
mediate (Dr. Keith Schmude, Ames, Iowa, Iowa State University
of Science and Technology, Department of Biochemistry and
Biophysics.
More pyridoxalchemy.
Private communication.
1963.)
Nonenzymic Transamination of Leucine with
5-Carboxylate Analogs of Pyridoxal
Attempts were made to determine the effects of the carboxylate group in aliphatic side chains in the ^-position of
pyridoxal on the nonenzymic transamination of leucine.
The
synthetic compounds, 5-"carboxypyridoxal" (CPL), with a carboxylate group substituted for the hydroxymethyl group of
pyridoxal, and "alpha^-pyridoxalylacetic acid (PLA), with a
propionate group substituted for the hydroxymethyl group.
229
were considered (Figures 1, 17). (As of this date, "alpha^pyrldoxalylformic acid," FPL or PLF, having an acetate group
substituted for the hydroxymethyl group, was not available in
free aldehyde form.)
"Carboxypyridoxal" vs. pyridoxal phosphate in acetate buffer
An estimate on the effectiveness of 5-"carboxypyridoxal"
as an amino group acceptor, compared with pyridoxal phosphate,
was made using a solution of about 0.01 M GPL, (not yet recrystallzed, since recrystallizatlon was a slow process).
The nonenzymlc transamination of leucine with pyridoxal phos­
phate proceeded about three times farther in the same time as
compared to that with "carboxypyridoxal", under the conditions
indicated (Table 27). Under similar conditions of pH and
buffer concentration the pseudo first-order rate constants
were 12.8 X 10~^/sec. for pyridoxal phosphate and 28 X 10"^/
sec. for deoxypyridoxal, as amino group acceptors.
The lesser reactivity of "carboxypyridoxal", as compared
to pyridoxal phosphate under these conditions, may be partly
due to formation of an internal lactal between the adjacent
carboxylate and formyl groups on the pyridine ring, especially
in an acidic media.
Difficulty experienced in,this laboratory
in preparing the next higher homolog of carboxypyridoxal
"alpha^-pyridoxalylformic acid," (PLP or PPL) has been at­
tributed to formation of an internal lactal structure (See
Figure 36).
230
Table 27.
Changes in milllmolar concentrations of pyridoxal
analog (PA) and ketoisocaproate (KIC) after 9.5
hours reaction^
+ mM KIC^
PLP
GPL
-mM PA
0.98
1.38°
+0.05
l.OOQ
0.37
+0.02
0.43°
0.3 (?)
^Initial conditions: 2 mM PA, 75 mM leu, 1.0 M acetate,
PH 3.9, 250
^Formation of KIC based on A ^^05 m
method and
calculated from standard curve
^ (Figure 8)
^Decrease in pyridoxal analog based on A A^QQ ^ by QH
method and calculation from standard curve
(Figure
8)
d
Decrease in pyridoxal phosphate based on
EOA method and calculation from standard
curve (Figure 8)
I
AA
by
^Decrease in carboxypyridoxal based on A A
^ by EGA
method and calculation from PLP standard
curve
(Figure 8); however, the spectrum of CPL in EOA had the
highest peak at 318 and not near 350 mp.; A_.g increased.
Instead of decreased as A„-„
^
350
5
"alpha -Pyridoxalylacetate" vs. deoxypyridoxal in imidazole
buffer
Bresnahan in this laboratory found that "alpha^-pyridoxalylacetate" (PLA) reacted 1.7 times faster than deoxy­
pyridoxal with leucine in 1.0 M imidazole buffer (pH 8.6,
231
14.5 X icT^/sec. for PLA, 8.5 X 10"^/sec. for DPI, lO"^ M
each with 0.02 M leu).
However, these reactions seemed to be
zero-order in that absorbancy was directly proportional
(linear) with time.
Comparisons of reactions in unbuffered solutions
General acid-base catalysis by buffers has been found to
interfere with observation of intramolecular acid-base cata­
lysis by phosphate in pyridoxal phosphate, because buffer
catalytic constants for nonenzymic transamination of leucine
with deoxypyridoxal are greater than those with pyridoxal
phosphate. For this reason measurement of effects of func­
tional groups in the 5-position of pyridoxal on the nonenzymic
transamination of leucine should be carried out in unbuffered
solutions (prepared by careful adjustment of pH).
The nonenzymic transamination of leucine with "alpha^pyrldoxalylacetic acid" (PLA) was compared to that with deoxy­
pyridoxal (DPL) and was found to be between 10 to 20 times
faster.
Calculation of relative rate constants were based on
AA
by the QH method (data forAA,...
was unsuitable
305 mju
400 m)i
due to the slow reaction of PLA with QH); and also on Inverse
times compared to the same AA
and the
for DPL, pH 7.2 and 9.2,
305 mjji
value at pH 8.1 of 3.6 X 10-7/sec. (Figure 20).
The average value of k
Ob s
calculated for the reaction with
PLA was 6 +1.7 X 10"G/sec. (Figure 31).
Direct spectral data were Inconsistent, perhaps due to
Figure 31.
Apparent zero-order kinetics for nonenzymlc
transaminations of leucine with pyrldoxal
phosphate (PLP) and with "alpha5-pyrldoxalylacetate" (PLA)& followed practically to
completion
PLP, pH 6.4
X 10^/sec.
A A vs. (KIC) by QH method
1.6 - 2.8^; 4.0°
OA vs. (PLP) by QH method
3.1 - 2.3^
OA vs. (aldlmlne) by direct
spectra
5.5
PLA, pH 8.1
A A vs. (KIC) by QH method
(zero-order only after 24
hrs.)
c
6 4^1.7
•A vs. (ketlmlne or PMA) by
direct spectra
2.0
#A vs. (aldlmlne) by direct
spectra
(not zero-order)
2.0
^Initial conditions;
unbuffered, 25°C
50 mM leu, 10 mM PLP or PLA,
^Calculated from Initial and final slopes In Guggen­
heim plots
^Estimated relative rate compared to reaction with DPL
at pH 6.8, under similar conditions, by taking a
ratio of inverse times to complete the same amount of
reaction; (k = k t /t )
X
y y X
233
or
zOH
25
75
TIME,HRS.
100
125
234
formation of white precipitate in the dark orange solution of
PLA and leucine after twenty four hours (about 25^) reaction.
Titration of the same preparation of PLA by Dr. Tomita in this
laboratory also resulted in the formation of a brown precipi­
tate above pH 9.5 which seemed to be a decomposition product.
As in several of the preceding nonenzymic transamination re­
actions with other pyridoxal analogs, under different con­
ditions, the data were not consistent with pseudo first-order
kinetics in that linear plots of absorbancy vs. time were
obtained, from both the QH and the spectral analyses.
QH method suggested slower initial rates.
The
The spectral
method, however, gave the same rates for decrease in absorb­
ancy at 405
as for increase in absorbancy at 330 mp
(Figure 31).
The significance and interpretation of apparent zeroorder kinetics in nonenzymic transamination systems under a
variety of conditions for these reactions followed practically
to completion are unclear to this author.
Banks et al.,
(1961) also found zero-order kinetics at high concentrations
of alanine in the reaction with pyridoxal, although the re­
action did follow first-order kinetics at relatively low con­
centrations of alanine.
Zero-order kinetics suggests the
saturation of a catalyst in the rate-limiting step in the re­
action, as in the case of enzyme kinetics.
Comparisons of the reactivities of these two carboxylate
analogs (PLA and GPL) were made with deoxypyridoxal and pyri-
235
doxal phosphate at pH 4.5 (near the acid pH optimum), where
PLA should be more stable than at high pH.
Each reaction was
followed for a period of one day with as many different ana­
lytical methods as possible. Direct spectral analysis in­
dicated that the reactions could best be followed by changes
in concentrations of the hydrogen-bonded aldimine, with de­
creasing absorbancies at 400 to 4l5 m;i, and of products
(probably pyridoxamine analogs or ketimines) with increasing
absorbancies at about 325 m^. Decreases in absorbancies at
400 to 415 mja were constant with time (lineaaf^lots of A vs.
t), so these slopes (-m) were compared as relative reaction
rates (Table 28). It was difficult to control the pH of the
unbuffered solutions.
Table 28.
Pyridoxal
analog
Spectral results comparing relative reactivities
of deoxypyridpxal (DPL), pyridoxal phosphate
(PLP), "alphax-pyridoxalylacetate" (PLA), and
"5-carboxypyridoxal" (CPL)&
Initial
pH
-m X lO^A/hr
pH after about
(X) days reaction
1
DPL
5.7
4.6
6.6 (2)
PLP
3.1
4.9
5.4 (1)
PLA
1.0
4.7
4.9 (1)
GPL
0
5.3
1
^Initial conditions: 0.10 M leu, 1.0 mM pyridoxal
analog, unbuffered, 25°C
236
The spectrum of leucine and "carboxypyridoxal" was not
significantly different from the spectrum of CPL alone, and
no changes in either spectrum were detected one day after
preparation of the solutions.
These spectra had absorption
maxima at 320 and 282 m/i. Indicating the absence of free
aldehyde.
Perhaps this preparation of "CPL" crystals was
primarily the lactal form, the 4-carboxylic acid or lactone
or some other form without a free 4-formyl group.
The
"carboxypyridoxal" sample had been recrystallized as bright
yellow crystals and dried in
vacuo.
The spectrum of "alpha^
pyridoxalylacetate" indicated the presence of about only ^
as much pyridoxal analog as expected from the amount weighed
out.
This sample had been recrystallized once from water,
in the same manner as the sample that had shown a fast re­
action at pH 8 (Figure 31).
The relative amounts of imine
formed were in the order of PIiA< PLP<DPL, under these con­
ditions.
Results of analyses by the QH method are indicated by
the extents of reaction in Table 29.
Also, no reactions were
detected with the carboxylate analogs of pyridoxal and ethanolamine. (See Experimental for details.)
Chromatography of reaction solutions
Products of transamination were also identified by thinlayer chromatography of reaction solutions of leucine and
pyridoxal analogs, other than deoxypyridoxal (Tables 30, 31).
237
Table 29. Extent of reaction of leucine with pyrldoxal ana­
logs measured by the QH method^
A
b
305
Relative
extent
c
A,
400
Relative
extent
Hours
DPL
0.038
1.0
0.024
1.0
18.5
PLP
0.142
3.7
0.054
2.3
18.25
PLA
0
0
0.008
0.3
18.33
CPL
0
0
0
0
18.0
^0.20 ml. aliquots of 1 mM In (PL analog) and (KIC) with
0.5 ml. 1 mM OH In 10 ml. 0.01 N HCl
^ Aonc
, measure of (KIC)(corrected for changes In
305 m}i
control sample (-leu)
° ^400 mjJi' measure of (PL analog) (corrected for changes
In control sample (-leu)
These results, in general, for the reactions of leucine with
pyrldoxal phosphate and with "alpha^-pyrldoxalylacetate"
agreed with those for the reaction of leucine with deoxypyrldoxal (Figures 18, 32). However,the pink ninhydrin spots of
pyridoxamine phosphate (PMP) and "alpha^-pyridoxaminoacetate"
(PMA) were more difficult to detect than the orange ninhydrin
spot of deoxypyridoxamine, because the former were located
just above the purple ninhydrin spot for leucine (See Rp
238
Table 30. Thin-layer silica gel chromatography of solutions
after about 200 hours reaction at 25°^
Rp values
Spot color
Dinitrophenylhydrazine-treated solutions developed with
n-BuOH (5), HOAc (l), HO (4) :
DNP-KIC
0.7
yellow
Additional spots after ninhydrin spray:
,
leu
0.3
purple
PMP
0.4
pink
PMA
0.4
pink
^Initial conditions; 50 mM leu, 10 mM PLP, pH 6.4, or
10 mM PLA, pH 8.1, unbuffered
values in Tables 30, 3l).
In view of the fact that the more quantitative analyt­
ical methods did not detect any reaction of GPL with leucine,
the appearance of dinitrophenylhydrazone of ketoisocaproate
(along with an unidentified yellow spot) on a chromatogram
of this mixture was quite puzzling (Table 31). Failure to
sometimes detect transamination products in the other re­
action solutions was attributed to insufficient sample size.
»
239
Table 31.
Thin-layer silica gel chromatography
solutions"
of reaction
Rp values and spot colors
Reagents:
Glbbs' + Ninhydrin
Spot
Sample
Standard
identity
PLP
4-
leu
DNPH + Ninhydrin
Sample
Standard
0.22-0.29 brown
with pink edges
PLP
0.31
brown
leu
KIC
0.15
0.36
yellow
0.31-0.32
pink
0.53
0.53
yellow
PMP (?)
(0.03)
brown
DNPH (2 spots)
PLA fleu
PLA
leu
0.26-0.30 brown
with pink edges
0.57
brown
0.74-0.83 0.77-0.81
yellow
0.57
0.50 (0.31) 0.49(0.28)
yellow
0.55 (0.27)
pink
KIC
yellow
DNPH (2)
GPL + leu
GPL
leu
0.76-0.83
0.74-0.87
yellow
0.69
0.73
yellow
0.69
0.73
yellow
0.37
pink
240
Table 31.
(continued)
R values and spot colors
F
Reagents:
Spot
Identity
KIC
Glbbs' +
Sample
Nlnhydrin
Standard
DNPH
+
Ninhydrin
Sample
Standard
0.43
0.46
yellow
?
0.24
yellow
DNPH (2)
0.78-0.84
0.80-0.87
^Developed with upoer layer of n-BuOH (5), HO (4), E
OH (1)
^
^Initial conditions: 0.10 M leu, 1.0 mM PLP, PLA or
CPL, pH 4.7 - 5.4, unbuffered
Nonenzymic Transaminations of Alanine Methyl
Ester with Pyridoxal Analogs
The slowness of nonenzymic transaminations of leucine
with pyridoxal analogs in unbuffered systems encouraged the
search for a more reactive amino group donor (than leucine).
Previous studies of nonenzymic transamination of amino acid
esters indicated that these amino group donors (in the
absence of metal ion catalysts) were at least as effective
Figure 32.
Thin-layer silica gel chromatography of reaction solutions of leucine
and deoxypyridoxal
Column no.
1
leu
2
DPL
3
KIC
4
DPL 4- leu near t ^
e
DPL
5
6
DPL + leu near t ^
e
Spots were treated with DNPH before development with n-BuOH (5), water (4), ETOH
(l) - upper layer and sprayed with ninhydrin after development
Rp values next to spots calculated for center of spot
Color code:
Y - yellow
Pr - purple
Or - orange
a
Initial conditions:
n
1.0 mM DPL, 0.01 M leu, 0.5 M acetate, pH 4.4, 25 C
^Same as a, except 0.05 M leu, pH 4.1
y ]qT2.
y 069
y ÎQ62
yQo^ôO
0
05r
q
yQopH
044
0o;
0,
00,
y"^Q32
QOÔ
1
1
2
qo3
3
4
243
as the corresponding amino acids in the presence of metal ion
catalysts.
Also, metal ions do not catalyze the nonenzymic
transamination of amino acid esters (Cennamo, 1964).
Alanine
methyl ester was chosen because of its greater solubility than
leucine esters, as well as leucine, itself. (Higher solubility
means a higher ratio of amino group donor to amino group
acceptor, resulting in a greater fraction of pyridoxal analog
converted to imine, and simplifying the kinetics by imposing
pseudo first-order conditions with respect to the pyridoxal
analog.)
Stoichiometry
Rates of decrease in concentrations of reactants (free
pyridoxal analog plus its imine of alanine methyl ester)
followed directly by decreasing absorbancies at 390 to 420
mfi were practically equal to rates of increase in concen­
trations of products (pyridoxamine analogs or ketimines),
followed directly by increasing absorbancies at about 325 nyi
(except in the cases of pyridoxal and pyridoxal methochloride,
which already had strong absorption in this region due to the
predominance of hemiacetal form).
Furthermore, a linear plot
of absorbancy at 327 mp against absorbancy at 398 m;i through­
out the time period of the reaction of 0.10 M alanine methyl
ester (AME) with 0.001 M deoxypyridoxal at pH 4.6 was obtained.
The formation of pyruvate methyl ester was not followed
by the quinolylhydrazone method because of the very slow
244
(about four hours compared to one hour for ketoisocaproate)
reaction of the keto acid ester with the QH reagent at room
temperature.
However, the spectrum of the quinolylhydrazone
of the keto acid ester had a sharp absorption maximum at 300
m^, which was very similar to that of QH products with free
keto acids.
Chromatography of reaction solutions
Evidence for products of nonenzymic transamination of
alanine methyl ester with deoxypyridoxal (DPL) and "alpha pyridoxalylacetate," (PLA) was obtained by chromatography of
of these reaction solutions on thin-layer silica gel with the
developing and spotting techniques previously described for
the leucine systems.
No such evidence was obtained from a
reaction solution of "5-carboxylpyridoxal" and AME (Figure
33).
Kinetics
Absorption spectral changes did not follow simple firstorder kinetics (although the conditions were pseudo firstorder, with respect to deoxypyridoxal), since the reaction
proceeded to an equilibrium point at greater than 30fo con­
version of deoxypyridoxal to deoxypyridoxamine, by an excess
of the amino group donor. However, Guggenheim plots were
linear and pseudo first-order rate constants were calculated,
despite decreases in pH of the unbuffered solutions during
the reactions (Table 32). The reactions were followed spec-
Figure 33.
Thin-layer silica gel chromatography of reaction solutions
Column no.
1
AME
2
DPL
3
DPL + AME at t
4
PLA
5
PLA + AME near t^ (after about 1 day reaction)
6
CPL
7
CPL + AME near t^
DNPH reagent added to each spot before development with n-BuOH (5),
water (4), EtOH (l)-upper layer; sprayed with ninhydrin after
development
Rp values next to spots calculated for center of spot
Color code:
Y - yellow
Pr - purple
Pk - pink
Or - orange
Initial conditions: 0.20 M alanine methyl ester (AME), 1.0 mM DPL, PLA or.
CPL, pH 7.5 unbuffered, 25^0
0,6
y 10.80
y)Qr9
y lori
OTI
Q73
062
0,41
PU
Pr
028-034
0,24
QI6
010
Pk\Q09
I
Pk
247
Table 32. Pseudo first-order rate constants (k
) for nonenzymic transaminations of alanine
methyl
ester (AME) with pyridoxal analogs in unbuffered
aqueous solutions, 25°C!®-
Initial conditions
Spectral
changes
followed at
wavelengths
(mp)
k
X
obs
lo5/sec.
pH after
t (hrs.)
No.
PH
(PL analog)
1
5.1
DPL
-415, -400,+325 10.4 +0.3
4.6
DPL
-398,+327 9.2 4^0.2
4.4 (13)
2
4.5
DPL
-415
12.4 +0.5
3.5 (24)
3
4.0
PLP
-420,-415,-400 16.9 +0.2
+325
3.85 (5)
4
4.9
PL
-405
5.6
3.6 (27)
5
4.5
PL
-400
3.6
3.5 (27)
6
4.5
PLM
-405
4.3
3.5 (27)
7
7.5
PL
-400
3.9
7.2 (3)
5.5 (22)
8
7.5
PLM
-405
9
6.8
PLP
-400 ^325 13.1 +0.5
6.5 (2.5)
10
7.2
PLA
-415 +335
2.4 +0.6
6.65 (6.5
11
7.2
GPL
-328,+275
negligible 5.6 (29)
25
4.2 (5)
7.2 (3)
5.5 (22)
^Initial conditions: 0.50 M AME, 1.0 M PL analog, except
for 0.10 M AÎ4E in reactionsland 0.5 mM PL in reaction 4
248
trally in about 5 to 30 hours to completion at 25°C.
Rates of reaction started at pH 4 to 5
The pseudo first-order rate constant for the reaction of
0.5 M alanine methyl ester (AME) with 1.0 mM deoxypyridoxal
(DPL) was not much greater than that for the reaction of 0.1
M AME with 1.0 mM DPL (Table 32; reactions 1_, 2). Pyridoxal
phosphate (PLP) is a better amino group acceptor than deoxy­
pyridoxal under similar conditions, since the rate constant
for the reaction of PLP with AME was about 1.3 to 1.4 times
greater than that for the reaction of DPL with AME (Table 32;
reactions 2,3). This may be further evidence for intramolec­
ular acid-base catalysis of aldimine to ketimine conversion
by the phosphate group of PLP, Pyridoxal (PL) and its Nmethyl analog (PLM) are poorer amino group acceptors under
similar conditions, since the rate constants of their re­
actions with AME are only about one-third as great as that
of the reaction of DPL with AME (Table 32; reactions 2,4-6).
This is further evidence that pyridoxal and analogs with 5hydroxymethyl groups have decreased reactivity toward amino
group donors, because their aldehyde groups are largely tied
up as internal or cyclic hemiacetals.
Rates of reactions started at neutral pH
The comparison of pseudo first-order rate constants for
the reactions of pyridoxal (PL) and pyridoxal methochloride
(PLM) with alanine methyl ester, which were started at pH 7 - 5 ,
249
was striking evidence that the aldlmine protonated at the
pyridine ring nitrogen atom (PLM) is much more reactive (by
a factor of 6 times under these conditions) than the aldlmine
which is unprotonated (PL) (Table 32; reactions 7, 8). How­
ever, the rate constant for the pyridoxal reactions did not
decrease significantly at this pH from that at the lower pH,
The rate constant for the reaction with PLP in neutral solu­
tion did decrease somewhat, as expected from dissociation of
the pyridinium group of the most reactive imine, even though
more imine was formed as expected from the formation constant
as a function of pH (Figure 15), and as evidenced by the 15
mp higher absorption band (Table 32, reaction 9).
The rate constant for the reaction of PLA with AME was
less than one-fifth as great as that of the reaction of PLP
with AME under these conditions.
The reaction of CPL with
AME was negligible after one day, indicating negligible
amounts of free aldehyde in the CPL preparation that was used.
It was likely also that the preparation of PLA was of
questionable purity, since greater reaction rates were ob­
served with PLA than with PLP and leucine. These prepara­
tions had been stored as aqueous stock solutions in the
frozen state, and these solutions were adjusted to about pH
8 to favor the free aldehyde form, by opening up the lactal
ring, after thawing and before adjusting the pH of the re­
action solution with AME to neutrality.
This procedure is
expected to be successful in yielding a high proportion of
250
reactive free aldehyde, if the preparation contains potential
aldehyde bound in the form of an internal lactal with the
adjacent carboxylate group.
In general, the rate constants for the reaction of pyri­
doxal with alanine methyl ester were only about one-third as
great as those obtained for the nonenzymic transamination of
L-alanine with pyridoxal at 100°C in unbuffered solutions
(Table 33; Blake jet
, 1963). They were about l|- times
greater than that of the reaction of pyridoxal with alanine
at 25°C (Banks et
, 1961), although the latter were meas­
ured at pH 10 in N,N-dimethylglycine buffer, which catalyzed
the reaction.
And they were about one-third as great as
those for the reaction of phenylglycine and pyridoxal, cata­
lyzed by 1.8 M imidazole at pH 8.6 but slightly higher than
that for the same reaction in 1.0 M imidazole (Bruice and
Topping, 1963a), and 15 times greater than those for the re­
action of alanine and pyridoxal, unbuffered at pH 8. (Fleck
and Alberty, 1962).
Furthermore, alanine methyl ester systems were more re­
active than leucine model transamination systems (Table 34).
The unbuffered system of the ester and pyridoxal had a pseudo
first-order rate constant of about 1§ times greater than that
of leucine and pyridoxal catalyzed by aluminum ion and acetate
buffer. The rate constant for the reaction of the ester with
pyridoxal phosphate was about 17 times greater than that for
leucine and PLP (both reaction solutions being unbuffered).
251
Table 33. Comparison of selected rate constants for nonenzymic transamination of pyridoxal with amino
acids under various conditions
Initial conditions
5
0.20 M L-ala
0.01 M PL
-1
k
X 10 sec."
obs
Reference
Blake et al., 1963
14.8
pH 4.15, unbuffered
100°c
+ 0.10 M HOAc + 0.10 M NaOAc
28.5
+ 5 X 10"5 M Alum
50
0.17 M ala
-4
10 • M PL
0.26
Fleck and Alberty,
1962
2.6
Banks et al., 1961
pH 8.0, 0.05 M NaOAc
25°C
0.2 M Ala
0.01 M PL
pH lOj 0.1 M Nj N- dimethylglycine
25°C
10"^ M = jz^gly = PL
14.7
pH 8.6, 1.8 M Im
10-3 M = p^gly = PL, 1.0 M Im 3.0
0.1 M Im 0.08
30OC
Bruice and Topping,
1963a ^
252
Table 33.
(continued)
Initial conditions
k , X 10^ sec. ^
obs
0.50 M Ala methyl ester
3.63.9
Reference
(Bresnahan and Albert,
1964)a
1.0 mM PL
pH 7.5 or 4.5, unbuffered
25°C
,
Most of these results with AI# (Tables 32 and 34) were
reported at the Seventh West Central States Biochemistry
Conference, November J, 1964, Iowa City Iowa, Sheryl
Bresnahan, Jerry Albert, and David E. Metzler. Nonenzymic transamination of alanine methyl ester and
leucine with pyridoxal analogs
Table 34.
Comparison of pseudo first-order rate constants
for nonenzymic transamination of alanine methyl
ester and of leucine with pyridoxal analogs
Leu
Initial Conditions
X lo5 sec. ^
k
obs 25OC
0.120 M Leu
10 mM PL
2.0 mM alum
pH 4.2, 0.1 M acetate
buffer
AME
Initial conditions
0.50 M AME
2.7
3.6-3.9
1.0 mM PL
pH 4.5, unbuffered
253
Table 34.
(continued)
Leu
Initial conditions
k
c
_2
X 10 sec.
AME
Initial conditions
Obs 25OC
0..10 M leu
0.50 M AI®
20 mM PLP
1.0 mM PLP
pH 3.9, unbuffered
0.95
16.9 +0.2
pH 4.0, unbuffered
0.05 M leu
10 mM PLP
pH 8.6, 1.0 M imidazole 0.25
pH 8.3, 1.0 M phosphate 0.14
0.10 M leu
0.50 M AME
10 mM DPL
1.0 mM DPL
pH 4.2, unbuffered
0.34
pH 4.2, 1,0 M acetate
buffer
2.8
pH 4.2, 1.0 M acetate
buffer and 1.0 mM
4.7
12.4 +0.5
pH 4.5, unbuffered
0.01 M leucine ethyl ester^
0.1 mM DPL
0.10 M phosphate, pH6.0
2.1-
1.0 M phosphate
4.3
5.2
^Robert Johnson, Ames, Iowa, Iowa State University of
Science and Technology, Department of Biochemistry and
Biophysics. Non-enzymatic transamination of leucine
ethyl ester with pyridoxal analogs. Private communi­
cation. 1964.
254
And the rate constant for the reaction of the ester with
deoxypyridoxal was about 37 times greater than that for leu­
cine and DPL (both reaction solutions being unbuffered). The
ester reaction in unbuffered solutions is still about 4 to 5
times greater than the leucine reaction in the presence of
1.0 M acetate buffer and about 2|- to 3 times greater than the
latter with an additional cuprlc ion catalyst.
The rate con­
stant for the reaction of DPL with AME was about 3 times
greater than that with leucine ethyl ester buffered with
phosphate (Table 34).
Alanine methyl ester has been probably one of the best
amino group donors (for which quantitative rate constants
have been obtained) for comparing reactivities of pyridoxal
analogs in unbuffered solutions at 25°C.
Not only were the
reactions fast enough to be practically followed to comple­
tion in 5 to 30 hours, but contaminating metal ions do not
Interfere (a problem in unbuffered amino acid systems).
Other advantages of the alanine methyl ester system include '
its high solubility, permitting measurements of pseudo firstorder rate constants (first-order with respect to small
amounts of pyridoxal analogs, which conserves the precious
synthetic compound), and the ease of measurement of the rate
constant by direct absorption spectral data (using cell
spacers, if necessary) with the Guggenheim method, without
apparently any side reactions or unexplainable stoichiometry
(which may be observed with the slower reacting amino acid
255
systems).
The enhanced reactivity of amino acid esters over free
amino acids as amino group donors in nonenzymic transamina­
tion systems has been explained on the basis of eliminating
the negative charge of the carboxyl group (Cennamo, 1964;
Guirard and Snell, 1964). Metal ion chelation is prevented
by loss of the carboxylate ligand, but inductive withdrawal
of electrons around the alpha-carbon in the amino acid
residue of the aldimine is enhanced by loss of the negative
charge on the carboxyl group in amino acid esters. It has
further been suggested that transaminases enhance the trans­
amination of amino acids in a similar manner, by blocking
the negative charge of the carboxyl group.
Increases and decreases in absorption maxima of imines
of pyridoxal phosphate and peptides, amino acid esters or
proteins were observed within about an hour after mixing
(Christensen, 1958). These spectral changes were incorrectly
attributed to formation of certain hydrated and ionic forms
of the imines. More recent evidence has been presented for
the assignment of these same absorption bands to the tauto­
meric ketoenamine and enolimine forms (A, B-Pigure 4; Review
of Pertinent Literature; Martell, 1963).
Although the spec­
tral-time changes observed by Christensen have been con­
firmed (William P. Jencks, Ithaca, New York, Cornell Univer­
sity, Department of Chemistry. Spectral-time changes for
Imines of pyridoxal phosphate and substituted amino acids.
256
Private communication. 1964), a satisfactory interpretation
has not been presented.
These findings cast some doubt on the reported nonenzymic
transaminations of pyridoxal analogs with glycine ethyl ester
(Tomita and Metzler, 1964), leucine esters (Johnson, 1964;
Cennamo, 1964), and with alanine methyl ester in this labor­
atory. However, in contradiction to this possibility, Christensen recovered pyridoxal completely by the ethanolimine
method after his spectral changes approached equilibrium,
whereas decreases of pyridoxal or of pyridoxal analogs were
followed by this same method (Johnson, 1964; Cennamo, 1964).
In addition, keto acid was detected in the reaction mixture
at 100°C by the dinitrophenylhydrazone method (Cennamo, 1964).
Products of nonenzymic transamination of pyridoxal analogs
and alanine methyl ester also were detected by thin-layer
chromatography and by the quinolylhydrazone method.
At present this author believes that, in general, either
nonenzymic transamination or at least ketimine formation re­
actions were followed, but that the spectral phenomena ob­
served by Christensen may be side reactions that might account
for the stoichiometric and kinetic anomalies observed during
some of these transamination reactions.
Ketimine formation
and the slow breakdown of ketimine to products could explain
loss of pyridine aldehyde, the spectral changes, and the
difficulty experienced in detecting transamination products
by thin-layer chromatography.
Perhaps, the ketimines of
257
amino acid esters and pyridoxal analogs are more stable than
those of free amino acids, because a free carboxylate group
catalyzes or enhances the hydrolysis of ketimine to products.
This would imply that imidazole had a similar effect on the
free carboxylate group of phenylglycine in its ketimine with
pyridoxal since an unusually slow hydrolysis to products was
reported (Bruice and Topping, 1963a).
Deuterium Isotope Effect on the Rate-Limiting
Step in Nonenzymic Transamination
The tri-deuterated isotope of leucine containing deu­
terium atoms instead of the protium atoms in the alpha-and
beta-positions (D^-leu) was compared with protio-leucine
(H^-leu) in two different model transamination systems.
With deoxypyridoxal as the amino group acceptor and imidazole
as the buffer and catalyst the reaction with H^-leucine was
about fifteen times faster than that with D^-leucine based
on the ratio of second-order rate constants, and about ten
times faster, based on the inverse ratio of times for 10^
decrease in deoxypyridoxal (Table 35). The reaction of D^leucine proceeded about
times faster in approximately 50^
DgO. Acid-base catalysis by D^O is suggested, since the
dissociation constant for D^O is less than that for HgO.
With pyridoxal phosphate as the amino group acceptor
and acetate as the buffer and catalyst the reaction with
leucine was about 8 times faster than that with D^-leucine
258
Table 35. Deuterium isotope effect in transamination of leu­
cine with deoxypyridoxal
Apparent second-order rate constants
Relative rates
X 10~^/M/sec.
k /k
H,-leu
^
17.5
_+l.l
15.3
Dq-leu (in H_0)
^
^
l.l4
+0.14
1.0
1.95
1.7
D -leu (in 50^
DgO)
+0.13
Time for
, % DPL lost/
% reaction hrs.
H,-leu
J
Dg-leu (in HgO)
D -leu (in 50^
3
DgO)
1.0 hr.
45^/8 hrs.
9.7
9.7 hrs.
l8^/24 hrs.
1.0
6.6 hrs.
25^/24 hrs.
1.4
^Initial conditions: 30 mI4 leu, 20 mM DPL, 1.0 M imi­
dazole, pH 6.8 - 7.0, 250
^Prom plots of fo decrease in deoxypyridoxal vs. time
in hours, followed by the ethanolimine method
(Table 36; Figure 34), based on the ratio of inverse times
to complete a certain percentage of the reaction between 26
and 38^. The ratio of rate constants or initial rates, sub­
stantiated by a Guggenheim plot of the data from the protlo-
259
Table 36. Deuterium isotope effect in transamination of leu­
cine with pyridoxal phosphate^
A) Decreases in (PLP) were followed by ethanolimine (EOA)
method and calculations were made from plots of A
vs. t hrs.; for Ho-leu, assumed A (A at t ) =
350 NYI
1.32 and A (A at t ) = 0.09; for
° D_-leu
assumed A^ - A - 1.23 and A % 1.34; then
chose
points:
^
°
For D^-leu
(A^-A^) % (A^-A^)
t hrs. (l/t„) ^ ^
Tvy
1.00
0.37
30.0
1.05
0.32
26.0
0.95
0.37
30.0
4.0
1.00
0.32
26.0
3.2
H
®
7.5
24
7.5
For H^-leu
B) Decreases in (PLP) were followed by QH method and calcu­
lations were made from plots of A^..
vs. t hrs.; for
H--leu, assumed A = 0.465, A =
'"H 0.105; for Do^ leu, assumed A_ - A =0. 36 and A^ = 0.450:
then chose points?
For D^-leu
A^OQ
(A^-A^) 0 (A^-A^)
t hrs.
0.315
0.135
37.5
7.5
0.340
0.110
30.5
0.330
0.135
37.5
4.0
0.355
0.110
30.5
2.8
24
8.6
For Hg-leu
^Initial conditions: 75 mM leu, 10 mI4 PLP, 1.0 M acetate
buffer, 250, pH 3.88 for H^-leu, pH 3.78 for D^-leu
26o
Table 36. (continued)
C) Formation of ketolsocaproate was followed by QH method
and calculations were made from plots of
vs. t
hrs.: for Ho-leu, assumed A = 0.120,
^^
A- = 0.475; for Do-leu, assumed A - A = 0.355 and A
®
J
o
e
^
= 0.115; then chose points:
For D^-leu A^q^
(A^-A^)
0.230
0.185
0.160
0.115
0.070
0.045
0.235
0.190
0.165
0.115
0.070
0.045
t hrs. k^/k
D
32.4
19.8
11.3
30
24
6
6.7
12
6
For H^-leu
32.4
19.8
11.3
4.5
2.0
1.0
Average k^/k^ = 8.0 +1.3
leucine reaction, was in agreement with the estimated relative
rates (Table 37).
This deuterium isotope effect on the rate of nonenzymic
transamination of leucine with pyridoxal analogs directly
confirmed the previous kinetic findings of others that the
rate-determining step involves the cleavage of the alphacarbon-hydrogen bond in the amino acid-aldimine of pyridoxal.
The over-all rate of transamination of glutamate was reduced,
as predicted by the binary mechanism and by model system
studies, in which the tautomerization of aldimine to ketimine
Figure 34. Deuterium isotope effect in transamination of
leucine with pyridoxal phosphate^
0
PLP^, decrease in (PLP) from reaction with H^-leu
followed by EOA method
D
PLP^j decrease in (PLP) from reaction with H^-leu
followed by QH method
A
KIC^, increase in (KIC) from H^-leu followed by QH
method
#
PLP^^, decrease in (PLP) from reaction with D^-leu
followed by EOA method
•I
PLP , decrease in (PLP) from reaction with Do-leu
D
followed by QH method
A
increase in (KIC) from D -leu followed by QH
p
J
method
^Reaction conditions: Table 36
262
i
TIME, HAS.
263
Table 37.
Comparison of Ho-leu and Do-leu based on cal­
culated rate
constants from data of Figure
34
k,
°
H^-leu
X lO^/seo.
from log (A' - A) vs. t
(Guggenheim plots
l6.6
350 mp
16.6
305
17.2
400
k = (dp/dt)V(0.01 M PLP) X
lO^/sec.
22
D^-leu
3.2
k„/k^
^
22/3.2 = 6.9
16.6/3.2 =5.2
^Best straight lines through points, discarding initial
and final points which "curve off"
^Initial rates obtained graphically from A^^g
vs. t
intermediate is the rate-determining step (Review of Pertinent
Literature), when alpha-deuterio-glutamate was a substrate for
glutamic-aspartic transaminase (Banks et al., 1963; Vernon,
1964).
Smaller isotope effects were observed by Blake et al.
(1963) for the nonenzymic transamination of L-deuterioalanine (containing deuterium atoms in the alpha-and-betapositions) with pyridoxal at lOO^C. However, the isotope
264
rate ratio k /k of 2,3 increased with an increase in catalyst
concentration (buffer or metal ion); e. g., k^/k^ was found to
be 3.5 in the presence of aluminum ion (Table 38).
General
acid-base catalysis by D^O was also suggested from the latter
results: k(H-ala in HgO)/k(H-ala in D^O) = 3 and k^k^ in
HgO = 2.2, but in D^O = 2.9.
Table 38. Deuterium isotope effects in transamination of Lalanine and pyridoxal, lOO^G (Blake et al., 1963)
Additional
conditions^
k]_ X lO^/sec.
k X loV^/sec.
/k
V D
unbuffered
1.48
7.40
2 . 2 +0.1
with 0.10 M rIOAc,
0.10 M NaOAc
2.85
14.25-
2.4 +0.06
with 5 X 10-5 M
alum.
5.0
25
3.5 +0.02
in DgO (990)
4.25
21.25
2.9 +0.02
a
Initial conditions: 0.01 M PL, 0.20 M L-ala, pH 4.15
The surprisingly large isotope effects observed in the
nonenzymic transamination of Ey-leucine with pyridoxal phos­
phate and with deoxypyridoxal in 1 M buffers are probably due
to general acid-base catalysis by the high concentrations of
acetate and imidazole buffers. The effect of buffers on the
265
Isotope ratio may be checked by varying the buffer concentra­
tion. The fact that the isotope effect was found to be two
times greater for deoxypyridoxal than for pyridoxal phosphate,
as the amino group acceptor, may be due to the greater buffer
catalytic constants found for deoxypyridoxal compared to pyri­
doxal phosphate.
Also, the nature of the buffer and the pH
differed in the two experiments.
Furthermore the isotope ratio could be increased by an
increase in concentration of aldimine according to Blake e;t
al. (1963).
This would qualitatively account for the dif­
ferences in the observed isotope ratios in the nonenzymic
transaminations of Dy-leucine with deoxypyridoxal and pyri­
doxal phosphate and of D^-alanine with pyridoxal, because
the imine formation constant for deoxypyridoxal and leucine
is about 2 to 3 times greater than that for pyridoxal and
leucine (Figure 15), and about two orders of magnitude (lOO
times) greater than that for pyridoxal and alanine.
A value of k^/k^ of 7.4 at 25°C is expected if only the
stretching frequencies of the carbon-protium and carbondeuterium bonds are considered: k^/k^ = e ( h t ^ - h f ^ ) / 2 k T ) .
This however, is only a simplification which ignores many
possible effects in solution other than the immediate bonds
which are broken (Levine, 1 9 6 3 ) .
"For acid-base reactions of hydrogen bound to carbon,
the ratio k^/k^ has always been found to be as large as 2.7
and sometimes as large as 10. Such large effects can be used
266
as evidence that a hydrogen atom transfer is rate-determining."
(Frost and Pearson, 1961, p. 344). The prediction of such
isotope rate constant ratios depends on the theoretical equa­
tion for the rate and on the approximations included.
For
the most exact equation, k /k = I/ 0 . 0 5 8 = 16.Y, at 2$°, in
H D
the gas phase, neglecting solvent effects. Furthermore,
large deuterium-isotope effects have been reported ranging
from 10 to 2 6 , for k /k (Melander, i 9 6 0 ) .
IT
D
Future Pyridoxal Analogs Possibly Capable of Enhanced
Intramolecular Catalysis of Nonenzymic Transamination
Nonenzymic transamination reactions between pyridoxal
phosphate and amino acids and between pyridoxamine phosphate
and keto acids are in progress (Vernon, 1964).
Another recent
report on studies of general-base catalysis in model trans­
amination reactions has revealed that the reactions of other
(simpler) pyridoxal analogs, such as 3-hydroxypyridine-4aldehyde, with amino acids are being measured in model systems
(Bruice, 1964). Although these studies with relatively
common or simple pyridoxal analogs are necessary to firmly
establish the characteristics of nonenzymic transamination
quantitatively, the author believes that the most exciting
approach to enzymic catalysis is the study of synthetic pyri­
doxal analogs with side chains having functional groups cap­
able of intramolecular acid-base catalysis of the intermediate
imines with amino acid esters.
267
Studies of such pyridoxal analogs depend upon their
successful syntheses.
Syntheses of the "classical" pyridoxal
analogs, such as deoxypyridoxal and pyridoxal phosphate, and
other vitamin Bg compounds-are outlined in Figure 35.
Pre­
sumably, the N-methyl analog of deoxypyridoxal, XIII, could
be synthesized also, in a simpler manner than the synthesis
of the N-methyl analog of pyridoxal (II), as depicted in
this Figure. From the model system studies it became apparent
that the effect of a positively-charged pyridine nitrogen atom
on the rate of nonenzymic transamination could best be ob­
served with the N-methyl analogs of deoxypyridoxal, XIII, or
of pyridoxal phosphate XII, than with pyridoxal-N-methochloride. These two N-methyl analogs, XII and XIII, would
not only be useful for study in model systems
but also would
be interesting for comparison of their spectra, their binding,
and activity with apoenzymes of pyridoxal phosphate-dependent
enzymes.
The 5-carboxylate analogs of pyridoxal, IX and XI, were
more difficult to obtain and keep as the free aldehyde forms,
probably due to internal or cyclic lactal formation (Figure
36).
It seems likely that these compounds would react best
with amino group donors in model transamination systems in
neutral or alkaline solution, since acidic conditions favor
lactal formation.
These reaction rates were sometimes higher
and sometimes lower than those for reactions with deoxypyri­
doxal and pyridoxal phosphate, depending upon the conditions
Figure 35. Syntheses of some pyridoxal analogs, indicating commercially avail­
able form (C):
XIV
I
pyridoxol (PO) (C-hydrochloride) from pyridoxamine (PM), VI (Cdihydrochloride), (Wagner and Polkers, 1964)
pyridoxal (PL) (C-hemiacetal hydrochloride) from pyridoxal (Brooks,
1950;
IV
5-deoxypyridoxal (DPL) from pyridoxal (Heyl et
, 1953), from pyridoxal (by Dr. Isao Tomita in this laboratoryJ, or from pyridoxamine,
via deoxypyridoxamine (DPM), VIII (not shown), (Kuroda, 1964)
V
pyridoxal-5-phosphate (PLP) (C-monohydrate) from pyridoxamine, via
pyridoxamine-5-phosphate (PMP), VII, (C-dihydrochloride), (Heyl et al.
1951; Peterson _et
, 1953; Peterson and Sober, 1954; Kuroda, 1^3sïJ
II
XII
XIII
pyridoxal-N-methochloride (PLM), (Brooks, 196O; Heyl e;t a^., 1951)
N-methopyridoxal-5-phosphate (MPLP), from pyridoxal (Heyl et al.,
1951)
'
5-deoxypyridoxal-N-methochloride (DPLM), from deoxypyridoxal
(proposed)
MnOg
O^OH^
H9
-OH
MONO
H
HCl
Ù CI®
; H Li
g.NHgOH
[H]
CH3OH
2.NQHC03
I.
HCl
^ CH3I
CHs
i+.ngci
®HQ3P0
I. Mn02'1B"
a
1^4^
I. NggCOj^ ocu- ^
2. 50Cl£(benzco'5
0
H Cl®
II
HG-fi'-OH
Hg/^d
'
H
CI 0
_
0
HONO
2.Mn02%"
HO ^ "^^^HgCl®
H
H3POuj.
IH
-Celi+e
pH4
g.BgCl
Figure 36.
IX
Syntheses of 5-carboxylate analogs of pyrldoxal from pyridoxol:
"5-carboxypyridoxal" (CPL) via an isopropylidene pyridoxol
X
"alpha^-pyrldoxalylformate" (FPL or PLF) via a 5-chloromethyl iso­
propylidene pyrldoxal (Tomita e^ a^. 1964)
XI
"alpha^-pyridoxalylacetate" (APL or PLA) via a 5-chloromethyliso­
propylidene pyrldoxal
(Isao Tomita, Ames, Iowa, Iowa State University of Science and
Technology, Department of Biochemistry and Biophysics. Syntheses
of 5-carboxylate analogs of pyrldoxal. Private communication.
1964).
OH
OH
OH
ZE A
^ V/
0
r-OH
l-NEUTR
'NEUTRALIZE
2. to]
0
socia/
Dânzeney
2.MnOpB
H Cle
o^
NcigCOg
2 CN®
^ '.H3O®
2. MnOg B
H Cl®
1.
NaOEf
,.CH2^0eE+),
E+Og OgEf
MnOp Q
nn
272
of pH and concentration and relative amount of free aldehyde
in the preparation of the carboxylate analog used.
More quan­
titative measurements await further purification of these ana­
logs.
These carboxylate analogs may be considered more valuable
as intermediates en route to pyridoxal analogs with substit­
uents at the ^-position capable of carrying out acid-base
catalysis of their intermediate imines with amino acids. For
example, amide linkages may be formed between the carboxylic
acid side chain and bases such as histamine, which would yield
a pyridoxal analog with an imidazole group at the end of this
side chain. Intermolecular, general acid-base catalysis of
nonenzymic transamination by imidazole buffer has been studied
and discussed.
An imidazole group in a side chain attached
to pyridoxal may be even more efficient as an intramolecular
acid-base catalyst of the nonenzymic transamination (Figure
37-B). Synthesis of such pyridoxal analogs may begin with
protection of the 4-formyl group, with ephedrine, which is
easily removed by a pH adjustment (Figure 37-A).
In this scheme, an ester linkage may also be formed with
the carboxylate side chain, and other acidic or basic func­
tional groups at the other end of these alcohol side chains
may be introduced such as the quanidino, amino, triazolic or
tetrazolic groups. Attention may have to be given to deter­
mining the appropriate atomic length of the side chain which
Figure 37.
Possible coenzyme analogs or model active sites of transamination of
amino acids; syntheses suggested starting with ephedrine as pro­
tective agent of the 4-formyl group of pyridoxal analogs with 5carboxylate side chains
His+ cxm me
9-
>
HO
pjn
_^H3
HNCH3
®NH,
H©
H©
-Ephedn'ne
0
..
k,
^PL)
ro
HN
NH&
OH
z=
OP-O
-NH
©
BDP-PLP
•O®
275
will be capable of acting upon the proton-transfer sites of
the imine with maximum catalytic enhancement.
Particular
side chains with such functional groups may give specific
catalytic activities to the corresponding pyridoxal ana­
logs.
Transaldimination may enhance the rate of the nonenzymic
transamination by activating the formyl carbon atom through
a preformed intramolecular aldimine.
An example of such a
pyridoxal analog may be model compound C-Figure 37•
By sub­
stituting transaldimination for imine formation between a
pyridoxal analog and an amino acid, the enzymic environment
in transaminases is approached even more closely.
A methyl substituent at the pyridine ring nitrogen atom
may be all that is necessary to maintain a strong positive
charge on the nitrogen atom and to increase its electrophilicity, and thus, enhancement of the catalysis of the
bond-breaking process in the imine.
All of these above as­
pects of intramolecular catalytic functions are incorporated
into model compound D-(Figure 37):
1) substituent on pyridine
ring nitrogen atom resulting in increased electrophilicity,
2) unsubstituted phenolate group, 3) activated formyl carbon
atom due to preformed intramolecular aldimine, requiring a
transaldimination reaction with an amino acid substrate, 4)
imidazole or other acid-base group with appropriate side chain
length to effect intramolecular catalysis of the reaction of
an amino acid imine intermediate.
276
A possible PLP coenzyme analog with a nucleotide "handle"
is the synthetic PLP-ADP compound of Kuroda (1963b). This
analog may be interesting to study in nonenzymic as well as
in enzymic systems. Such pyridoxal analogs suggested here
may also prove to be potent inhibitors of PLP-dependent
enzymes and may possess pharmacological activités.
277
I
SUMMARY OF CONCLUSIONS
Certain pyridoxal analogs have been studied in aqueous
solution, and the extents and rates of their reactions with
amino acids have been measured. Rate constants for these nonenzymic transamination reactions are compared and correlated
with results reported in the literature.
The absorption spectrum of pyridoxal-N-methochloride at
pH values near 4 was found to represent a mixture of two
é
ionic forms, and measurements of the absorbancies at 293-5
and 323 mp at several pH values near 4 permitted a precise
evaluation of the apparent pK
&
value of 4.05 +0.03 for
dissociation of the phenolic group.
This was compared to the
corresponding value of 4.20 for pyridoxal, suggesting the
effect of N-methylation on the acidity of this group.
The equilibrium constant for the formation of the
hydrogen-bonded aldimine of this pyridoxal analog with valine
was quantitatively measured and the log
' value of -0.02
+_0.1 was found. This was compared to the corresponding value
of -O.27 for pyridoxal.
Similarly, relatively less imine was
formed from pyridoxal methochloride and leucine than from
pyridoxal and leucine in neutral or alkaline solution. The
explanation is that a positive charge on the pyridine ring
nitrogen atom inhibits resonance structures with a positive
charge in positions on the ring needed to stabilize the polar­
ization of the carbonyl group, since the formyl carbon atom
278
is the nucleophillc site of attack by the amino group of an
amino acid.
Also, the rates of nonenzymic transamination of leucine
and alanine methyl ester with pyridoxal methochloride were
greater than those with pyridoxal, especially at pH values
above the pKSi of the pyridinium group for the pyridoxal imines.
These findings are consistent with the reactive intermediate
having a positive charge on the pyridine ring nitrogen atom
as a result of N-methylation. The pyridoxal methochloride
imine was more unstable than the pyridoxal imine in that the
positive charge on the pyridine ring nitrogen atom enhanced
aidimine to ketimine tautomerization, since the quinoid
structure intermediate to the aldimine and ketimine forms is •
more favorable by resonance stabilization in the case of
pyridoxal methochloride than of pyridoxal.
The nonenzymic transamination of leucine with 5-deoxypyridoxal was studied stoichiometrically and kinetically;
deuxypyridoxal was considered
a reference compound to
which other pyridoxal analogs with functional groups In the
5-position may be compared. Products of the nonenzymic trans­
amination were Identified by thin-layer chromatography.
Ketoisocaproate and deoxypyridoxal were quantitatively meas­
ured by a modified method of analysis using quinolylhydrazine. Deoxypyridoxal, as well as pyridoxal, pyridoxal metho­
chloride, and pyridoxal phosphate were also quantitatively
measured in portions of reaction solutions by the ethanoli-
279
mine method.
The pH profile for the rate in the acid range,
in the absence of added catalysts, and for the imine for­
mation constants of leucine and deoxypyridoxal, also suggested
that the reactive intermediate has a positive charge on the
pyridine ring nitrogen atom as a result of N-protonation.
The phosphate group of pyridoxal phosphate enhanced the
nonenzymic transamination of leucine between 5 and 11 times,
compared with the methyl group of deoxypyridoxal, in the acid
pH range in the absence of added buffer. From the pH-rate
profile of this reaction the catalytic constant for the
monobasic form of this phosphate group was calculated to be
twice as great as that of the dibasic form. However, intermolecular catalysis by phosphate buffer as a function of pH
indicated that almost all of the catalytic constant for this
buffer could be attributed to the monobasic form.
The hydroxymethyl group at position 5 of pyridoxal had
a negative effect on nonenzymic transamination rates because
of hemiacetal formation blocking the essential formyl group.
Substitution of this group with carboxylate side chains, such
as propionate, indicated rate enhancement, possibly as a
result of Intramolecular acid-base catalysis.
Intramolecular effects by substituents in the 5-position
of pyridoxal analog,s in catalysis of nonenzymic transamination
were probably best demonstrated in unbuffered solutions, with
a good amino group donor, such as alanine methyl ester.
Blocking of the negative charge of the carboxyl group of an
280
amino acid, as In alanine methyl ester, led to rapid re­
actions with pyridoxal analogs.
These nonenzymic transamination reactions were also
catalyzed by acetate buffer.
The catalytic constant for
acetic acid was calculated to be about twice as great as
that for acetate. Imidazole buffer was also found to be
a good general acid-base catalyst.
However, metal ions,
such as copper (ll) and aluminum (ill), were found to be
good intramolecular catalysts, compared to zinc (ll), as
chelates of the imlnes in forming more of these inter­
mediates in the nonenzymic transamination, and thus re­
latively small amounts were as effective as much larger
amounts of buffer.
Finally, the deuterium i'sotope effect in the non­
enzymic transamination of deuterio-leucine, (under dif­
ferent conditions, average k /k was about 8), confirmed
n D
the rate-determining step to be the breaking of the alphahydrogen-carbon bond of the pyridoxylidene amino acid.
281
LITERATURE CITED
AlbertJ A. and J. N. Phillips
1956
Ionization constants of heterocyclic substances.
II. Hydroxyderivatlves of nitrogenous slxmembered ring-compounds. Chemical Society
(London) Journal 1956: 1294-1304.
Anderson, P. J. and A. E. Martell
1964 Pyridoxal phosphate: molecular species in solu­
tion. American Chemical Society Journal 86:
715-720.
Banks, 3.E.G., A. A. Diamantls, and C. A. Vernon
1961
Transamination. II. The non-enzymic reactions
between pyridoxamlne and pyruvic acid and between
pyridoxal and alanine. Chemical Society (London)
Journal 196I: 4235-4247.
Banks, B.E.C., A. J. Lawrence, C. A. Vernon, and J. F. Wootton
1963
Kinetic studies of glutamic-aspartic transaminase
(pig heart muscle). International Union of Bio- '
chemistry Symposium Series 3O: 197-215.
Blake, M. I., P. P. Siegel, J. J. Katz, and M. Kilpatrick
1963 Deuterium isotope effects in transamination: Lalanine and pyridoxal. American Chemical Society
Journal 85: 294-297.
Block, R. J., E. L. Durrum, and G. Zweig
1955
A manual of paper chromatography and paper elec­
trophoresis. New York, New York, Academic Press,
Inc.
Bonavita, V. and V. Scardi
1959
Studies on glutamic-oxalacetic transaminase. II.
The properties of two derivatives of pyridoxal-5phosphate. Archives of Biochemistry and Biophysics
82: 300-309.
Braunstein, A. E.
1957
Pyridoxal phosphate. In P. D. Boyer, H. Lardy,
and K. Myrback, eds. The enzymes. 2nd ed. Vol.
2. pp. 113-184. New York, New York, Academic
Press, Inc.
282
Braunstein, A. E.
i960
Pyridoxal phosphate. In P. D. Boyer, H. Lardy,
. and K. Myrback, eds. The enzymes. 2nd ed. Vol.
2. pp. 113-184. New York, New York, Academic
Press, Inc.
Braunstein, A. E., und M. G. Krltzmann
1937 II. Die Gleichgewichts reaktion zwischen 1 ( )
Glutaminsaure und Brenztraubensaure, bzw. 1 ( )
Alanin und alpha-Keto glutarsaure. Enzymologia
2: 138-146.
Braunstein, A, E., and M, M. Shemyakin
1952
A theory of amino acid metabolism processes cata­
lyzed by pyridoxal phosphate enzymes. (Trans­
lated title) Akademiia Nauk SSSR Doklady 85:
1115-1118.
Braunstein, A. E. and M. M. Shemyakin
1953
A theory of amino acid metabolism processes cata­
lyzed by pyridoxal phosphate enzymes. (Trans­
lated title) Biokhimiia I8: 393-411.
Brooks, H. G., Jr.
i960
Vitamin Bg derivatives and related compounds.
Unpublished Ph.D. thesis. Ames, Iowa, Library,
Iowa State University of Science and Technology.
Bruice, T. C.
1964 Studies of general-base catalysis in model trans­
amination reactions. International Union of Bio­
chemistry Symposium Series 32: 265-266.
Bruice, T. C. and T. C. French
1964 Rates and equilibrium constants of imine formation
with pyridine-4-aldehyde and various amino acids.
Biochemical and Biophysical Research Communica­
tions 15: 402-408.
Bruice, T. C.and R. M. Topping
1962
The imidazole-catalyzed (non-metal ion mediated)
transamination of phenylglycine by pyridoxal:
a, reaction occurring at ambient temperatures by
way of Michaelis-Menten kinetics. American
Chemical Society Journal 84: 2448-2450.
Bruice, T. C. and R. M. Topping
1963a Catalytic reactions involving azomethines. I.
The imidazole catalysis of the transamination of
pyridoxal by alpha-aminophenylacetic acid.
American Chemical Society Journal 85: l480-l485.
283
Brulce, T. C. and R. M. Topping
1963b Catalytic reactions involving azomethines. II.
pH dependence of the Imidazole catalysis of trans­
amination of pyridoxal by alpha-amlnophenylacetic
acid. American Chemical Society Journal 85:
1486-1492.
Bruice, T. C. and R. M. Topping
1963c Catalytic reactions involving azomethines. III.
Influence of morpholine upon the imidazole cata­
lysis of transamination of pyridoxal by alphaaminophenylacetic acid; transamination of the
morpholine imine of pyridoxal. American Chemical
Society Journal 85: 1493-1496.
Bruice, T. C. and R. M. Topping
1963d The imidazole-imidazolium ion concerted general
acid, general base catalysis of the transamination
of pyridoxal by alpha-aminophenylacetic acid.
International Union of Biochemistry Symposium
Series 30: 29-56.
Cattaneo, J., J. C. Senez, et P. Beaumont
1960 Transamination non-enzymatiques du pyridoxal
phosphate et du pyridoxamine phosphate a'temper­
ature physiologique et en presence de métaux.
Biochimica et Biophysica Acta 44: 543-553Cennamo, C.
1954
Non-enzymatic transamination between peptides and
pyridoxal. Naturwissenschaften 4l: 39.
Cennamo, C.
1958 Sui meccanismo della ^transaminazione non enzymatLca
del peptidi. Societa Italiana di Biologia Sperimentale Bollettino 34: 214-217.
CennamoJ C.
1961
Transaminazione non enzimatica tra piridossale e
esteri di aminoacidi. Societa' Italiana di Bio­
logia Sperimentale Bollettino 37: 183-I87.
Cennamo, C.
19d2 Rlcerche sui composti Intermedl delle reazloni
non enzlmatiche tra aminoacidi e piridossale.
Societa Italiana di Biologia Sperimentale
Bollettino 38: 1037-1040.
Cennamo, C.
1963 Intermediate compounds of non-enzymic transamina­
tions. International Union of Biochemistry
Symposium Series 30: 83-90.
284
Cennamo, C.
1964
Metal ion-Independent non-enzymic transamination
between pyridoxal and amino acid esters, (to be
published in Biochimica et Biophysica Acta Nov.,
1964.)
Christensen, H. N.
1958
Three Schiff base types formed by amino acids,
peptides and proteins with pyridoxal and pyridoxal-5-phosphate. American Chemical Society
Journal 8O: 99-105.
Christensen, H. N.
1959
Cupric chelates of pyridoxylvaline and pyridoxylidenevaline. American Chemical Society Journal
81: 6495-6498.
Cordes, E. H. and W. P. Jencks
1962a General acid catalysis of semicarbazone forma­
tion. American Chemical Society Journal 84:
4319-4328.
Cordes, E. H. and W. P. Jencks
1962b Semicarbazone formation from pyridoxal, pyridoxal
phosphate and their Schiff bases. Biochemistry
1: 773-778.
Cordes, E. H. and W. P. Jencks
1963 The mechanism of hydrolysis of Schiff bases
derived from aliphatic amines. American Chemical
Society Journal 85: 2843-2848.
Dancis, J., J. Hutzler, and M. Levitz
1963 Thin-layer chromatography and spectrophotometry
of alpha-keto acid hydrazones. Biochimica et
Biophysica Acta 78: 85-90.
Davis, L., F. Roddy, and D. E. Metzler
1961
Metal chelates of imines derived from pyridoxal
and amino acids. American Chemical Society
Journal 83: 127-134.
Dempsey, ¥. B. and H. N. Christensen
1962 The specific binding of pyridoxal 5'-phosphate to
bovine plasma albumin. Journal of Biological
Chemistry 237: 1113-1120.
Eichorn, G. L. and J. W. Dawes
1954 The metal complexes of vitamin Bg and Schiffs
base derivatives. American Chemical Society
Journal 76: 5663-5667
285
Evangelopoulous, A. E. and I. W. Sizer
1963
Pig heart glutanilc-aspartic transaminase. Mech­
anism of transamination. National Academy of
Sciences Proceedings 49: 638-643.
Pasella, P., G. G. HammeSj and B. L. Vallee
1962 Concerning the role of metals in enzymic trans­
amination. Biochimica et Biophysica Acta 65:
142-143.
Pasella, P.^ H. Lis, N. Siliprandi, and C. Baglioni
1957 Electrophoretic and chromatographic study of some
chemical transamination reactions involving
vitamin B^. Biochimica et Biophysica Acta 23:
417-428.
Pasella, P., H. Lis, N. Siliprandi, and C. Baglioni
1958
On the formation of intermediate chelate compounds
in model reactions involving vitamin Bg.
Journal of Inorganic and Nuclear Chemistry 8:
620-624.
Fleck, G. M. and R. A. Alberty
1962
Kinetics of the reaction of pyridoxal and alanine.
Journal of Physical Chemistry 66: 1678-I682.
Frost, A. A. and R. G. Pearson
1961 Kinetics and mechanism. 2nd ed.
York, John Wiley and Sons, Inc.
New York, New
Gonnard, P.
1962
Acquisitions recentes sur le role coenzymatlque
de la vitamin Bg. Cosmétologie 45: 19-23.
Gonnard, P.
1963
Sur les hydrazines de phospho-5-pyridoxal, role
coenzymatlque et stabilité. International Union
of Biochemistry Symposium Series 30: 307-312.
Gonnard, P., J. Duhalt, M. Gamier, C. Nguyen-Philippon, and
N. Biogne
1964
Action of phospho-5'-pyridoximinotriazole on pyri• doxalo-enzymes. Biochimica et Biophysica Acta 8I:
548-559.
Gregerman, R. I. and H. N. Christensen
1956
Enzymatic and non-enzymatic dehydrochlorination
of ^ - chloro - L - alanine. American Chemical
Society Journal 220: 765-774.
286
GuggenheimJ E. A.
1926
On the determination of the velocity constant of
a unimolecular reaction. Philosophical Magazine
2: 538-543.
Guirard, B. M. and E. E. Snell
1964
Vitamin 3g function in transamination and decar­
boxylation reactions. In M. Plorkin and E. H.
Stotz, eds. Comprehensive biochemistry: grouptransfer reactions. Vol. 15. pp. 138-174. New
York, New York, American Elsevier Publishing Co.,
Inc.
Gustafson, R. I. and A. E. Martell
1957 Stabilities of metal chelates of pyridoxamine.
Archives of Biochemistry and Biophysics 68:
485-498.
Gyorgy, P,
1934
Vitamin Bp and the pellagra-like dermatitis in
rats. Nature (London) 133: 498-499.
Hammes, G. G. and P. Pasella
1963a The mechanism of enzymatic transamination. Inter­
national Union of Biochemistry Symposium Series
30: 185-195.
Hammes, G. G. and P. Fasella
1963b The interaction of glutamic-aspartic transaminase
with Dseudo substrates. American Chemical Society
Journal 85: 3929-3932.
Harbury, H. A. and K. A. Foley
1958
Molecular interaction of isoalloxazine derivatives.
National Academy of Sciences Proceedings 44; 662-
668.
Heinert, D. and A. E. Martell
1962
Pyridoxine and pyridoxal analogs. V. Syntheses
and infrared spectra of Schiff bases. American
Chemical Society Journal 84: 3257-3263.
Heinert, D. and A. E. Martell
1963a Pyridoxine and pyridoxal analogs. VI. Electron
absorption spectra of Schiff bases. American
Chemical Society Journal 85 : I83-I88.
Heinert, D. and A. E. Martell
1963b Pyridoxine and pyridoxal analogs. VII. Acid-base
equilibria of Schiff bases. American Chemical
Society Journal 85: 188-193.
287
Heinert, D. and A. E. Martell
1963c Pyridoxine and pyridoxal analogs. VIII.
Synthesis and infrared spectra of metal chelates.
American Chemical Society Journal 85: 1334-1337.
Henson, C. P. and ¥. W. Cleland
1964
Kinetic studies of glutamic oxalacetic trans­
aminase isozymes. Biochemistry 3: 338-34$.
Herbst, R. M.
1944 The transamination reaction.
Enzymology 4: 75-97.
Advances in
Heyl, D., S. A. Harris, and K. Folkers
1953
Chemistry of vitamin 3g. IX. Derivatives of 5deoxypyridoxine. American Chemical Society
Journal 75: 653-665.
Heyl, D., E. Luz, 8. A. Harris, and K. Folkers
1951
Phosphates of the vitamin Bg group. I. The
structure of codecarboxylase. American Chemical
Society Journal 73: 3430-3433.
Ikawa, M. and E. E. Snell
1954a Benzene analogs of pyridoxal: the reactions of
4-nitrosalicylaldehyde with amino acids. American
Chemical Society Journal 76: 653-665.
Ikawa, M. and E. E. Snell
1954b Oxidative deamination of amino acids by pyridoxal
and metal salts. American Chemical Society Journal
76: 4900-4902.
Irvin, J. L. and E. M. Irvin
1947 Spectrophotometric and potentiometric evaluation
of apparent acid dissociation exponents of various
4-aminoquinolines. American Chemical Society
Journal '69: IO9I-IO99.
Isenberg, I. and A. Szent-Gyorgi
1958 Free radical formation in riboflavin complexes.
National Academy of Sciences Proceedings 44:
857-862.
Jencks, ¥. P.
1963
Mechanism of enzyme action.
Biochemistry 32: 648-650.
Annual Review of
288
Jencks, W. P. and E. Cordes
1963 Transaldimlnation reactions of pyridoxal and
related compounds. International Union of Bio­
chemistry Symposium Series 30: 57-67.
Jenkins, W. T.
Glutamic-aspartic transaminase. V. The reaction
1961
with L-alanine. Journal of Biological Chemistry
236: 474-478.
Jenkins, W. T.
1963 Binary complexes involved in enzyraic transamina­
tion. International Union of Biochemistry
Symposium Series 30: 139-148.
Jenkins, ¥. T. and I. W. Sizer
i960 Glutamic-aspartic transaminase. IV. The mech­
anism of transamination. Journal of Biological
Chemistry 235: 620-624.
Jenkins, ¥. T., D. A. Yphantis, and I. W. Sizer
1959
Glutamic-aspartic transaminase. I. Assay,
purification, and general properties. Journal
of Biological Chemistry 234: 51-57.
Johnson, R.
1964 A study of non-enzymatic transamination reactions
of amino acid esters with pyridoxal and its ana­
logs. Unpublished paper presented to Iowa Academy
of Sciences, April, 1964. Ames, Iowa, Department
of Biochemistry and Biophysics, Iowa State Univer­
sity of Science and Technology.
Johnston, C. C., H. G. Brooks, J. D. Albert, and D. E.
Metzler
1963
Reactions of pyridoxal-N-methochloride with amino
acids. International Union of Biochemistry
Symposium Series 30: 69-81.
Junk, G. A. and H. J. Svec
1964 Deuterium exchange in the pyridoxal leucine system.
Journal of Organic Chemistry 29: 944-946.
Kalyankar, G. D. and E. E. Snell
1957 Differentiation of alpha-amino-acids and amines
by non-enzymatic transamination on paper chromatogramSi Nature (London) I8O: 1069-I070.
289
Ketelaar, J.A.A., C. van de Stolpe, and H. R. Gersmann
1951
Spectrophotometric studies of the solvation of
iodine in dioxan solution. Recueil des Travaux
Chimiques des Pays-Bas 71: 499-508.
Ketelaar, J.A.A.j C. van de Stolpe, A. Goudsmit, and W.
Dzcubas
1952 Spectrophotometric study of the solvation of
iodine in dioxan solution. [llj| Recueil des
Travaux Chimiques des Pays-Bas 7I: 1104-1114.
King, H. K.
1963 Studies on leucine decarboxylase. International
Union of Biochemistry Symposium Series 30: 253266.
King, H, K. and N. Lucas
1959
Interaction between pyridoxal phosphate and amino
acids. Biochemical Journal J2: 18P.
Kuroda, T.
1963a Synthetic studies of vitamin Bg derivatives. [l]
Synthesis of pyridoxal-5-phospnate and its related
compounds. Vitamins 28: 211-220.
Kuroda, T.
1963b Synthetic studies of vitamin Bg derivatives. [ll]|
Synthesis of adenosine di phosphate of vitamin Bg
group. Vitamins 28: 354-361.
Kuroda, T.
Synthetic studies of vitamin Bg derivatives.
1964
[ill]
Synthesis of 5-deoxy-derivatives of
vitamin Bg group. Vitamins 29; II6-II8.
Levine, L. H.
1963
Some aspects of L-amino acid oxidase catalysis.
Unpublished Ph.D. thesis. Ames, Iowa, Library,
Iowa State University of Science and Technology.
Lis, H., P. Pasella, C. Turano, and P. Vecchini
I90O
On the mechanism of action of glutamic-aspartic
transaminase: intermediate steps in the reaction,
Biochimica et Biophysica Acta 45: 529-536.
Longenecker, J. 3. and E. E. Snell
1957 The comparative activities of metal ions in
promoting pyridoxal-catalyzed reactions of amino
acids. American Chemical Society Journal 79: 142145.
290
Lucas, N.J H. K. King, and S. J. Brown
1962
Substrate attachment In enzymes: the Interaction
of pyrldoxal phosphate with amino acids. Bio­
chemical Journal 84: 118-124.
Lunn, A. K. and R. A. Morton
1952
Ultra-violet absorption spectra of pyridoxine and
related compounds. Analyst 77î 718-731.
Mackay, D.
1962 The mechanism of the reaction of cysteine with
pyrldoxal 5'-phosphate. Archives of Biochemistry
and Biophysics 99: 93-100.
Maklno, K. Y. Ooi, M. Matsudo, and T. Kuroda
1963 Some notes on the coenzyme activity of phosphopyrldoxal derivatives for the glutamic decar­
boxylase and glutamic aspartic transaminase.
International Union of Biochemistry Symposium
Series 30: 291-306.
Makino, K.
T. Kuroda
1962
Y. Ooi, M. Matsuda, M. Tsuji, M. Matsumoto, and
Some notes on the coenzyme activity of phosphopyrldoxal derivatives for the brain glutamic
decarboxylase. Biochemical and Biophysical Re­
search Communications 9: 246-251.
Martell, A. E.
1963
Schlff bases of pyrldoxal analogs: molecular
species in solution. International Union of Bio­
chemistry Symposium Series 30: 13-28.
Martin, R. B.
1964
Reactions of carbonyl compounds with amines and
derivatives. Journal of Physical Chemistry 68:
1369-1377.
Marvel, C. S. and N. Tarkoy
1957 Heat stability studies on chelates from Schlff
bases of salicylaldehyde derivatives. American
Chemical Society Journal 79: 6000-6002.
Mason, S. P.
1958
The tautomerlsm of N-hetero-aromatic hydroxycompounds. III. Ionisation constants. Chemical
Society (London) Journal 1958: 674-685.
291
Matsuo, Y.
1957a Formation of Schlff bases of pyridoxal phosphate:
reaction with metal Ions. American Chemical
Society Journal 79: 2011-2015.
Matsuo, Y.
1957b Pyridoxal catalysis of non-enzymatic transamina­
tion In ethanol solution. American Chemical
Society Journal 79: 2016-2019.
Melster, A.
1953 Sodium alpha-ketolsocaproate. In E. E. Snell, ed,
Biochemical preparations. Vol. 3. pp. 66-70.
New York, New York, John Wiley and Sons, Inc.
Melster, A.
1962
Amino grouptransfer. In P. D. Boyer, H, Lardy,
and K. Myrback, eds. The enzymes. 2nd ed.
Vol. 6. pp. 193-217. New York, New York,
Academic Press, Inc.
Melster, A. and P. A. Abendschein
1956
Chromatography of alpha-keto acid 2,4-dinltrophenylhydrazones and their hydrogénation products.
Analytical Chemistry 28: 171-173.
Melander, L. C.
i960
Isotope effects on reaction rates.
York, Ronald Press Company.
New York, New
Metzler, D. E.
1957 Equilibria between pyridoxal and amino acids and
their imlnes. American Chemical Journal 79:
485-490.
Metzler, D. E., M. Ikawa, and E. E. Snell
1954a A general mechanism for vitamin Bg-catalyzed re­
actions. American Chemical Society Journal 76:
648-652.
Metzler, D. E., J. 011vard,.and E. E. Snell
1954b Transamination of pyridoxamine and amino acids
with glyoxyllc acid. American Chemical Society
Journal 76: 644-648.
Metzler, D. E. and E. E. Snell
1952a Deamlnation of serine. I. Catalytic deamination
of serine and cysteine by pyridoxal and metal
salts. Journal of Biological Chemistry 198: 353361.
292
Metzler, D. E. and E. E. Snell
1952b Some transamination reactions Involving vitamin
Bg. American Chemical Society Journal 74: 979983.
Metzler, D. E. and E, E. Snell
1955
Spectra and Ionization constants of the vitamin B^
group and related 3-hydroxypyrldlne derivatives.
American Chemical Society Journal 77: 2431-2437.
Nakamoto, K. and A. E. Martell
1959a Pyrldoxlne and pyrldoxal analogs. Ill, Ultra­
violet absorption studies and solution equilibria
of 2- and 4-hydroxyinethyl-3-hydroxypyridines and
pyridine-2,3- and 4-aldehydes. American Chemical
Society Journal 8I: 5857-5863.
Nakamoto, K. and A. E. Martell
1959b Pyrldoxlne and pyrldoxal analogs. IV. Ultra­
violet spectra and solution equilibria of 3methoxypyrldlne-2 (and 4-)-aldehydes and of 3hydroxypyridine-2 (and 4-j-aldehydes. American
Chemical Society Journal ol: 5863-5869.
Neish, W.J.P.
1957
Alpha-Keto acid determinations. In D. Click, ed.
Methods of biochemical analysis. Vol. 5. pp.
107-179. New York, Mew York, Interscience
Publishers.
Olivard, J. and E. E. Snell
1955a Growth and enzymatic activities of vitamin Bg
analogues I. D-Alanlne synthesis. Journal
of Biological Chemistry 213: 203-214.
Olivard, J. and E. E. Snell
1955b Growth and enzjrmatic activities of vitamin Bg
analogues II. Synthesis of miscellaneous amino
acids. Journal of Biological Chemistry 213:
215-228.
Olivo, P., C. S. Rossi, and N. Slliprandi
1963
Non-enzymlc transaminations of gamma-amlnobutyrlc
acid. International Union of Biochemistry Sympos­
ium Series 30: 9I-IOI.
Ooi, Y.
1964
Coenzymatic activity of pyrldoxal phosphate de­
rivatives on glutamic-oxaloacetic transaminase
Vitamins 29: IOI-IO8.
293
Perault, A. M., B. Pullman, and C. Valdemoro
1951 Electronic aspects of the reactions of pyrldoxal
phosphate enzymes. Blochlmlca et Blophyslca Acta
46: 555-575.
Peterson, E, A. and H. A, Sober
1954 Preparation of crystalline phosphorylated deriv­
atives of vitamin Bg. American Chemical Society
Journal 76: 169-175.
Peterson, E, A., H. A. Sober, and A. Melster
1953 Pyrldoxamine phosphate; pyrldoxal phosphate. In
E. E. Snell, ed. Biochemical preparations. Vol.
3. pp. 29-39. New York, Nev? York, John Wiley
and Sons, Inc.
Pullman, B.
1963 A quantum-mechanical Investigation of pyridoxaldependent reactions. International Union of
Biochemistry Symposium Series 30: 103-121.
Rablnowltz, J. C. and E. E. Snell
1947 The vitamin IBg group. XII. Microbiological
activity and natural occurrence of pyrldoxamine
phosphate. Journal of Biological Chemistry 169:
643-650.
Robins, E., N. R. Roberts, K. M. Eydt, 0. H. Lowry, and D. E.
Smith
1956 . Mlcrodeterminatlon of alpha-keto acids with
special reference to malic, lactic, and glutamic
dehydrogenases in brain. Journal of Biological
Chemistry 2l8: 897-909.
Roze, U.
1964
The reaction between pyrldoxal phosphate and
cycloserine. Unpublished Ph.D. thesis. Saint
Louis, Missouri, Library, Washington University.
Scardi, V., e. V. Bonavita
1957 Determinazlone spettrofotometrlca del plrldossal
-5-fosfato. Socleta' Itallana dl Biologia
Sperimentale Bollettino 33: 1701-1702.
Slegel, F. P. and M. I. Blake
1962 Spectrophotometrlc determination of pyrldoxal in
the presence of pyrldoxamine and pyridoxlne.
Analytical Chemistry 34: 397-398.
294
SlzePj I. W. and W. T. Jenkins
1963 The reactivity of the pyrldoxal phosphate group
of glutamic aspartic aminotransferase. Inter­
national Union of Biochemistry Symposium.Series
30: 123-138.
Snell, E. E.
1944 The vitamin activities of "pyrldoxal"and "pyridoxamlne". Journal of Biological Chemistry 15^:
313-314.
Snell, E. E.
1945a The vitamin Bg group. IV. Evidence for the
occurrence of pyridoxamine and pyrldoxal in
natural products. Journal of Biological Chem­
istry 157: 491-505.
Snell, E. E.
1945b The vitamin Bg group. V. The reversible interconversion of pyrldoxal and pyridoxamine by
transamination reactions. American Chemical
Society Journal 67: 194-197.
Snell, E. E.
1958
Chemical structure in relation to biological
activities of vitamin Bg. In R. S. Harris, G.
F. Marrian, and K. V. Thlmann, eds. Vitamins
and hormones. Vol. 16. pp. 77-125. New York,
New York, Academic Press, Inc.
Snell, E. E.
1962 A comparison between some pyrldoxal-dependent
enzymatic and non-enzymatic reactions. Brookhaven
Symposia in Biology 15: 32-51.
Snell, E. E.
1963
Non-enzymatic reactions of pyrldoxal and their
significance. International Union of Biochemistry
Symposium Series 3O: 1-12.
Snell, E. E., P. M. Fasella, A. E. Braunsteln, and A. RosslPanelli, eds.
1963
Chemical and biological aspects of pyrldoxal
catalysis. International Union of Biochemistry
Symposium Series 30; 1-599.
Snell, E. E., and W. T. Jenkins
1959
The mechanism of the transamination reaction.
Journal of Cellular and Comparative Physiology
54, Supplement 1: 161-I77.
295
Snell, E. E. and D. E. Metzler
1956
Water soluble vitamins. II. Annual Review ofBiochemistry 25: 435-443.
Snell, E. E. and A. N. Rannefeld
1945
The vitamin
group. III. The vitamin activity
of pyridoxal and pyridoxamine for various organ­
isms. Journal of Biological Chemistry 157: 475489.
Spikner, J. E. and J, C, Towne
1952 Pluorometric microdetermination of alpha-keto
acids. Analytical Chemistry 34: l468-l471.
Tomita, I., H. G. Brooks, Jr., and D. E. Metzler
1964 Synthesis of vitamin Bg derivatives. II. 3Hydroxy-4-(hydroxymethyl)-2-methyl-5-pyridine
acetic acid and related substances. Unpublished
manuscript submitted to Journal of Medicinal
Chemistry. Ames, Iowa, Department of Biochem­
istry and Biophysics, Iowa State University of
Science and Technology.
Tomita, I. and D. E. Metzler
1964 A new analogue of pyridoxal phosphate;3-hydroxy2-methyl-5-propionlc acid-4-pyridlne aldehyde.
Unpublished paper presented at the l48th national
meeting of the American Chemical Society, Chicago,
Illinois, September 1964. Ames, Iowa, Depart­
ment of Biochemistry and Biophysics, Iowa State
University of Science and Technology.
Umbreit, W. W., D. J. O'Kane, and I. C. Gunsalus
1948 Function of the vitamin Bg group: mechanism of
transamination. Journal of Biological Chemistry
176: 629-637.
Velick, S. P. and J. Vavra
1962a Glutamic-oxalacetate transaminase. In P. D. Boyer,
H. Lardy, and K. Myrback, eds. The enzymes. 2nd
ed. Vol. 6. pp. 219-246. New York, New York,
Academic Press, Inc.
Velick, S. P-, and J. Vavra
1962b A kinetic and equilibrium analysis of the glutamic
oxaloacetate transaminase mechanism. Journal of
Biological Chemistry 237: 2109-2122.
296
VernonJ G. A.
1964
Glutamic-aspartlc transaminase (pig heart muscle).
A comparison of the enzymlc and non-enzymlc re­
actions. International Union of Biochemistry
Symposium Series 32: 259-260.
von Vlscontlnl, M,, C, Ebnother, und P. Karrer
1951
Konstltutlon und Synthese der Codecarboxylase,
Helvetica Chimica Acta 34: 1834-1842.
Wada, H. and E. E. Snell
1962 Enzymatic transamination of pyrldoxamlne I. With
oxaloacetate and alpha-ketoglutarate. Journal
of Biological Chemistry 237: 127-132.
Wagner, A. F. and K. Polkers
1964 Vitamins and coenzymes. New York, New York,
Interscience Publishers.
Westheimer, P. H.
i960 Enzyme models. In P. D. Boyer, H. Lardy, and
K. Myrback, eds. The enzymes. 2nd ed. Vol. 1.
pp. 261-268. New York, New York, Academic Press,
Inc.
Wiegand, R. G.
1956
The formation of pyridoxal and pyrldoxal 5-phos­
phate hydrazones. American Chemical Society
Journal 78: 5307-5309.
Williams, V. R. and J. B. Neilands
1954 Apparent ionization constants, spectral properties
and metal chelation of the cotranamlnases and
related compounds. Archives of Biochemistry and
Biophysics 53: 56-70.
Willstadter, E., T. A. Hamor, and J. L. Hoard
1963 Stereochemistry of the manganese (ll) - pyrldoxylidenevallne chelate. American Chemical
Society Journal 85: 1205-1206.
297
ACKNOWLEDGEMENTS
The author acknowledges the technical help and exchange
of information related to this research from a number of
persons, especially Miss Erika Rommel, Miss Louise Hodgin,
Mr. Donald DePrenger, Dr. Isao Tomita, Miss Sheryl Breshahan,
Mr. Robert Johnson, Dr. Keith Schmude, and Mr. F. Scott
Furbish.
Previous to this work on model reactions of pyridoxal
analogs for the past two years,the author was struggling with
the nature of the active site of.threonine dehydrase from
sheep liver.
One aspect of this enzyme work was included in
a publication from this laboratory, and several other pre­
liminary findings have since been confirmed and reported by
Dr. Willie McLemore.
However, this research project was not
given up by the author until he was convinced that results
from investigation of the active site of this enzyme might be
difficult to obtain unless further purification was accompHdied.
The author thanks Dr. Donald J. Graves for his interest and
helpful discussions in this respect.
The author also con­
gratulates Dr. Willie McLemore for his accomplishments in
this enzyme project, which the author has followed with
interest during his graduate studies.
The author is grateful to Dr. David E. Metzler, his major
professor, for helping to make the change in research proj­
ects with not too much loss in time. The constructive crit­
298
icisms, patience, and guidance which Dr. Metzler provided not
only during the laboratory work, but especially during the
writing of this thesis, have established a background of
training which will be a valuable asset to the author through­
out his scientific career.
The author has appreciated and depended on the moral
support and encouragement of his family, relatives, teachers,
and friends, in addition to that of Judy. Financial support
of graduate assistantships through Iowa State University's
Agricultural Experiment Station is gratefully acknowledged.
299
DEDICATION
This thesis is humbly dedicated to the praise and glory
of the living God who has sustained the author in times of
trial and forgiven him in times of temptation throughout his
life, and Who has made all things possible and continues to
make all things work together for good to those who love
Him. The witness of many scientists dedicated to Jesus
Christ has convinced the author that any apparent conflicts,
past, present, or future, between science and Christianity
are really only imaginary or at least paradoxical and are
due either to poor science, inadequate religion, or both.
For the understanding that what people can learn through
God's Word and through God's world is complementary, not
contradictory, the author especially acknowledges the witness
of Dr. Prank L.Lambert, Professor of Chemistry, Occidental
College; Dr. Ray Pahien, Professor of Chemical Engineering,
University of Florida; Dr. John A. Effenberger, Senior
Physical Chemist, DuPont Research Laboratories, Wilmington;
Pastor W. J. Fields, Memorial Lutheran Church of Ames; Dr.
Walter R. Hearn, Professor of Biochemistry, Iowa State Uni­
versity, and the American Scientific Affiliation.
300
VITA
The author was born to Mildred Lydla (Volssem) and Leo
Paul Albert in Milwaukee, Wisconsin, June 6, 1937. He
attended General Douglas MacArthur and Morgandale elementary
schools. He had many interests and his particular interest
in science did not emerge until his high school years. He
was presented the Bausch and Lomb Science Award upon gradua­
tion from Pulaski High School in 1955jranking third in a
class of about 415.
After his family moved to California he majored in chem­
istry at Occidental College, Los Angeles.
He started his
laboratory experience in organic chemistry in a NSF research
project, studying halogenation of organic compounds, directed
by Dr. Frank Lambert. He was active in the Beta Mu chapter
of Alpha Chi Sigma, the national professional chemistry
fraternity, and in the Theta Chapter of Kappa Mu Epsilon,
the national honorary mathematics society.
Industrial
'
experience during summer vacations included syntheses of
organic phosphorus compounds, ultraviolet and Infrared spec­
trophotometry, and inorganic analytical chemistry. He re­
ceived the American Institute of Chemists Award upon gradua­
tion with a Bachelor of Arts degree from Occidental College
in 1959J ranking 26th in a class of 287.
During his graduate studies in biochemistry at Iowa State
301
University he also had minors in organic chemistry and cell
biology. He became a candidate for the Ph.D. degree in 1961,
and was married to Judith J. Brown at the end of that year.
The committee in charge of his Ph.D. program included:
chairman: Dr. David E. Metzler, Professor of Biochemistry
major
: Dr. R. Scott Allen, Professor of Biochemistry
and Animal Science
1 st minor, organic chemistry: Dr. Glen A. Russell,
Professor of Chemistry
2 nd minor, cell biology: Dr. L. Evans Roth, Professor
of Biophysics and Assistant
Dean, Graduate College
member-at-large: Dr. Charlotte E. Roderuck, Professor
of Pood and Nutrition
He became a member of Phi Lambda Upsilon, the national
honorary chemistry society, through the Theta chapter at Iowa
State in 19^3. He is also a member of the American Chemical
Society and the American Scientific Affiliation. His bio­
chemical interests have been the mechanisms and functions of
naturally-occurring compounds in living cells.
Farewell to Iowa I
The author hopes that he will not be
considered a former lowan by the local mass media like many
others have been so branded for merely passing through this
weather-beaten and provincial land.